Pearson

Higher Nationals in
Computing
Unit: 6 Planning a Computing
Project

1 Thanoj E127966
Higher Nationals – Summative Assignment Feedback Form

Student Name/ID
Unit Title Unit: 6 Planning a Computing Project

Assignment Number 1 Assessor Ms Sumudu Samarakoon

Date Received 1st

Submission Date
submission
Date Received 2nd
Re-submission Date
submission

2 Thanoj E127966
Assessor Feedback:

LO1. Conduct small-scale research, information gathering and data collection to generate knowledge on an

identified subject

Pass, Merit & Distinction P1 P2 M1 D1

Descripts
LO2. Explore the features and business requirements of organisations in an identified sector

Pass, Merit & Distinction P3 P4 M2

Descripts

LO3. Produce project plans based on research of the chosen theme for an identified organisation

Pass, Merit & Distinction P5 M3

Descripts

LO4. Present your project recommendations and justifications of decisions made, based on research of the
identified theme and sector
Pass, Merit & Distinction P6 P7 P8 M4 D2
Descripts

* Please note that grade decisions are provisional. They are only confirmed once internal and external moderation has taken place and
grades decisions have been agreed at the assessment board.

Assessor Feedback:

Grade: Assessor Signature: Date:

Resubmission Feedback:

• Please note resubmission feedback is focussed only on the resubmitted work

Grade: Assessor Signature: Date:

3 Thanoj E127966
Internal Verifier’s Comments:

Signature & Date:

• Please note that grade decisions are provisional. They are only confirmed once internal and external
moderation has taken place and grades decisions have been agreed at the assessment board.

4 Thanoj E127966
Important Points:

1. It is strictly prohibited to use textboxes to add texts in the assignments, except for the
compulsory information. eg: Figures, tables of comparison etc. Adding text boxes in the body
except for the before mentioned compulsory information will result in rejection of your work.
2. Avoid using page borders in your assignment body.
3. Carefully check the hand in date and the instructions given in the assignment. Late submissions
will not be accepted.
4. Ensure that you give yourself enough time to complete the assignment by the due date.
5. Excuses of any nature will not be accepted for failure to hand in the work on time.
6. You must take responsibility for managing your own time effectively.
7. If you are unable to hand in your assignment on time and have valid reasons such as illness,
you may apply (in writing) for an extension.
8. Failure to achieve at least PASS criteria will result in a REFERRAL grade .
9. Non-submission of work without valid reasons will lead to an automatic RE FERRAL. You
will then be asked to complete an alternative assignment.
10. If you use other people’s work or ideas in your assignment, reference them properly using
HARVARD referencing system to avoid plagiarism. You have to provide both in-text citation
and a reference list.
11. If you are proven to be guilty of plagiarism or any academic misconduct, your grade could be
reduced to A REFERRAL or at worst you could be expelled from the course
12. Use word processing application spell check and grammar check function to help editing your
assignment.
13. Use footer function in the word processor to insert Your Name, Subject, Assignment No,
and Page Number on each page. This is useful if individual sheets become detached for any
reason.

5 Thanoj E127966
 STUDENT ASSESSMENT SUBMISSION AND DECLARATION

When submitting evidence for assessment, each student must sign a declaration confirming that the
work is their own.

Student name: H.M Thanoj (E127966) Assessor name: Ms Sumudu Samarakoon

Issue date: Submission date: Submitted on:

Programme:
BTEC HND in Computing

Unit: 6 Planning a Computing Project

Assignment number and title:

1 | Managing a research project on practical applications and limitations of the use of AI

Plagiarism
Plagiarism is a particular form of cheating. Plagiarism must be avoided at all costs and students who
break the rules, however innocently, may be penalised. It is your responsibility to ensure that you
understand correct referencing practices. As a university level student, you are expected to use
appropriate references throughout and keep carefully detailed notes of all your sources of materials for
material you have used in your work, including any material downloaded from the Internet. Please
consult the relevant unit lecturer or your course tutor if you need any further advice.

Guidelines for incorporating AI-generated content into assignments:

The use of AI-generated tools to enhance intellectual development is permitted; nevertheless, submitted
work must be original. It is not acceptable to pass off AI-generated work as your own

Student Declaration

Student declaration
I certify that the assignment submission is entirely my own work and I fully understand the consequences of
plagiarism. I understand that making a false declaration is a form of malpractice.

Student signature: Date:

6 Thanoj E127966
Assignment Brief

Student Name/ID H.M Thanoj (E12796)

Number

Unit Number and Title Unit: 6 Planning a Computing Project

Academic Year 2024/2025

Unit Tutor

Assignment Title Managing a research project on practical applications and

limitations of the use of AI

Issue Date
Submission Date

Submission Format
This assignment should be submitted to the ELMS as a single document on or before the
given submission date and time.

The document should comprise of

• A comprehensive project plan - including a work, time and resource allocation/ breakdown using
appropriate tools. A business area analysis Including the features and operational areas of the business
and the role of stakeholders and their impact on the success of the business.
• A research paper - including application and evaluation of quantitative and qualitative research
methods to generate relevant primary data and examination of secondary sources to collect relevant
secondary data and information.
• An Action plan – including recommendations and evaluation of project outcomes comparing the
decisions given in the project plan.

The assignment should be word-processed, and figures can be either hand drawn or
generated from a tool.
There is no specific word limit for this assignment.

7 Thanoj E127966
Unit Learning Outcomes
LO1 Conduct small-scale research, information gathering and data collection to generate knowledge on an
identified subject
LO2 Explore the features and business requirements of organisations in an identified sector.
LO3 Produce project plans based on research of the chosen theme for an identified organisation
LO4 Present your project recommendations and justifications of decisions made, based on research of the
identified theme and sector

Transferable skills and competencies developed

This unit offer students an opportunity to demonstrate the skills required for managing and
implementing a project.
The students will undertake independent research and investigation for carrying out and
executing a computing project which meets appropriate business aims and objectives.
Students will develop competencies of:
• project management skills
• research skills
• writing, investigative, field study, performance skills, as appropriate presentation skills.

8 Thanoj E127966
Assignment activity and guidance :

Research Theme: Artificial Intelligence

Research Topic: Practical applications and limitations of the use of AI in an identified sector
(e.g. e-commerce, healthcare, biomedical science, education, lifestyle, automotive
engineering, agriculture, gaming, finance)

“Artificial intelligence is at the forefront of innovation within Computer Science that uses a combination of logic,
algorithms and large data sets to produce an AI model. The AI model is created to perform specific tasks or make
predictions on supplied sets of input data, for example identifying patterns in weather data, internet search data or
analysis of medical data. Artificial intelligence is predicted to generate a potential impact to the global economy
of $13 – $15 Trillion by 2030, with sales of AI related hardware, software and services predicted to see a global
revenue of $900 billion. It is predicted that AI will boost the GDP of China by a little over 26% by 2039, and of
North America by 14.5%. AI requires the input of structured and labelled data, where the outputs are already
known. The input data sets to the AI model are intrinsically linked to study field to which the AI engine is to be
applied. The AI model can then be used to identify and recognise patterns and relationships within the input data.
This identification step is referred to as ‘training’ the AI model. Once this training is completed, the model can
then be used to make predictions and identify patterns within brand new data sets. This new data set can then be
added to the existing data set, so that the AI model keeps ‘growing’. As the model data set keeps expanding, and
the AI algorithms are modified and refined, this gives the impression that the AI is ‘learning’ and demonstrating
‘intelligence’. AI has been used extensively to analyse and process large and complex datasets produced by big
data systems, often in real time and using Computer Vision to extract data from image sources. Developing
Artificial Intelligence required a range of knowledge and skills across a broad range of computer science
disciplines. AI developers need to be familiar with the algorithms and techniques in fields such as machine
learning, natural language processing, computer vision and data science. Knowing the required computing skills
will help organisations recruit the correct resources to help develop and extend AI systems. Artificial Intelligence
has a range of benefits across many industry sectors. In the finance industry, AI is rapidly becoming a game
changer, using advanced algorithms, models and machine learning to carry out predictive analytics on large,
rapidly changing financial datasets to provide more accurate financial predictions. In the field of business
operations, AI automation is helping to support and enhance labour productivity, leading to greater cost savings
and increased efficiency. AI is also revolutionising the way businesses interact with their customers, by providing
AI driven expert systems to help customers resolve queries as well as providing personalised recommendations
based on customer choices and preferences. In the field of biomedical science, AI models help in the development
of new drug treatments for a range of diseases by searching and processing large scale medical and DNA datasets.
While Artificial Intelligence has numerous benefits in the analysis and processing of large data sets to solve

9 Thanoj E127966
problems, there are some clear risks to the application technology. AI systems respond to the data fed into the
model, and so if this data is not representative of the problem area under study, there is a likelihood that the output
of the AI model will be biased. In addition, there are security and privacy concerns on the source and storage of
the large datasets used for AI. The rise of the Deepfake image and the manipulation of the human voice is also a
concern because of the spread of misinformation. The wide-ranging effects of these risks mean that they can only
be dealt with by a diverse range of stakeholders, including computer scientists, law makers, governments, and
industry leaders. There are also incidental risks of AI in business, for example the increasing adoption of AI based
systems may increase unemployment across a range of sectors and workforce demographics. Learners will have
the opportunity to investigate the applications, benefits and limitations of Artificial Intelligence while exploring
the responsibilities and solutions to the problems it is being used to solve.”. (Pearson, 2024)

You are expected to carry out small-scale research in order to explore the practical applications and
limitations of the use of AI in an identified sector (e.g. e-commerce, healthcare, biomedical science,
education, lifestyle, automotive engineering, agriculture, gaming, finance). The research that you carry out
can be based on an organization that you have access to gather sufficient information to investigate the practical
applications and limitations of the use of AI.

The findings of the research should be presented in a professionally compiled report and the report should cover
the given tasks including,

• A comprehensive project plan - including a work, time and resource allocation/ breakdown using
appropriate tools. A business area analysis Including the features and operational areas of the business and
the role of stakeholders and their impact on the success of the business.
• A research paper - including application and evaluation of quantitative and qualitative research methods
to generate relevant primary data and examination of secondary sources to collect relevant secondary data
and information.
• An Action plan – including recommendations and evaluation of project outcomes comparing the decisions
given in the project plan.

Activity 1 - A comprehensive project plan

Activity 1
1.1. Select an organization of your choice that allows you to explore and study the relevant data.
Provide an introduction and background to your project and the chosen organization. Define the scope and
devise aims /objectives of the project that you are going to carry out. You also should include risks and
benefits of exploring the practical applications and limitations of the use of AI of the chosen organization/s
or the field.

10 Thanoj E127966
2.1. Plan a small-scale research project to explore practical applications and limitations of the use of AI.
Produce a comprehensive project plan including the following.
• Cost, scope, time, quality, communication, risk, and resources management plan.
• Comprehensive Work Breakdown Structure (WBS) with clearly defined activities and milestones.
• Gantt chart to illustrate project schedule with realistic time allocated for each activity and clearly defined
deadlines for milestones.

Activity 02 – Research Report

2.1. Discuss the features and operational areas of the chosen organization/s, the role and the impact of
stakeholders for the success of the business. You also need analyse the challenges the organization/s may
face in achieving the success and meeting business objectives by applying AI technologies to achieve
operational efficiency.

2.2. Carry out the research to investigate the “Practical applications and limitations of the use of AI in an
identified sector “ and generate relevant primary data by applying appropriate qualitative and quantitative
research methods. You need to examine secondary sources to collect relevant secondary data and information
to support the research. You then need to analyse the data and information and interpret the findings to
generate knowledge on how the application of AI supports business requirements in the identified
organization/s .

Activity 03: Action Plan

4.1 Communicate appropriate project recommendations derived from the research data analyzed for
technical and non-technical audiences and assess the extent to which the project recommendations
meet the needs of the chosen organization/s.
4.2 Discuss the reliability, accuracy, and the appropriateness of the research methods applied while
arguing and evaluating the planning recommendations made in the project plan comparing them to
the actual outcomes and the need of the chosen organization/s.

Recommended Resources
Please note that the resources listed are examples for you to use as a starting point in your research
– the list is not definitive.

11 Thanoj E127966
➢ Costley, C., Elliot, G. and Gibbs, P. (2010) Doing Work Based Research: Approaches to Enquiry for
Insider-researchers. London: SAGE.
➢ Dawson, C. (2016) Projects in Computing and Information Systems: A Student's Guide. UK: Pearson
Education.
➢ Gray, D. (2009) Doing Research in the Real World. 2nd Ed. London: SAGE.
➢ Guay, M., Schreiber, D. and Briones, S. (2016) The Ultimate Guide to Project Management: Learn
everything you need to successfully manage projects and get them done. Free Kindle Edition. US: Zapier
Inc.
➢ Lock, D. (2013) Project Management 8th Edition. UK: Routledge.
➢ Pinto, J.K. (2015) Project Management: Achieving Competitive Advantage 4th Ed. Pearson.

12 Thanoj E127966
Learning Outcomes and Assessment Criteria

Pass Merit Distinction

LO1 Conduct small-scale research, information gathering

and data collection to generate knowledge on an identified
subject

P1 Demonstrate qualitative
M1 Analyse data and D1 Interpret findings to
and quantitative research
information from primary generate knowledge on how
methods to generate relevant and secondary sources to the research theme supports
primary data for an identified generate knowledge on an business requirements in the
theme. identified theme. identified sector.
P2 Examine secondary
sources to collect relevant
secondary data and
information for an identified
theme.

LO2 Explore the features and business requirements of

organisations in an identified sector

P3 Discuss the features and

M2 Analyse the challenges
operational areas of a
to the success of a business
businesses in an identified in an identified sector.
sector.
P4 Discuss the role of
stakeholders and their
impact on the success of a
business.

13 Thanoj E127966
Pass Merit Distinction

LO3 Produce project plans based on research of the chosen

theme for an identified organisation

P5 Devise comprehensive
M3 Produce comprehensive D2 Evaluate the project
project plans for a chosen
project plans that effectively planning recommendations
scenario, including a work
consider aims, objectives and made in relation to the
and resource allocation
risks/benefits for an identified needs of the identified
breakdown using
organization. organisation and the
appropriate tools.
accuracy and reliability of
the research carried out.

LO4 Present your project recommendations and justifications

of decisions made, based on research of the identified theme
and sector

P6 Communicate M4 Assess the extent to which

appropriate project the project recommendations
recommendations for meet the needs of the identified
technical and nontechnical organisation, including
fullysupported rationales for
audiences.
planning decisions made.
P7 Present arguments for the
planning decisions made
when developing the project
plans.
P8 Discuss accuracy and
reliability of the different
research methods applied.

************************ END *************************

14 Thanoj E127966
Table of Contents
Plagiarism .............................................................................................................................................. 6
Activity 1, ............................................................................................................................................. 18
1.1 Introduction to AI in E-Commerce, ......................................................................................... 18
1.1.1 How AI is Revolutionizing E-Commerce? .................................................................................. 18
1.1.2 Introduction to Daraz, .................................................................................................................. 20
1.1.3 Daraz Sri lanka Company Structure............................................................................................. 21
1.1.4 Aim of the Project ........................................................................................................................ 26
1.1.5 Scope of the project ..................................................................................................................... 27
1.1.6 Impact of AI concept on E-Commerce sector and Daraz ............................................................ 28
1.1.7 Risk and Benefits ......................................................................................................................... 30
1.1.8 Objective of the Project................................................................................................................ 31
1.1.9 Purpose of the Project .................................................................................................................. 32
1.1.10 Significant Aim of this project, .................................................................................................. 34
1.2 Project Management Concept .................................................................................................. 35
1.2.1 Project Lifecycle .......................................................................................................................... 40
1.2.2 Project Life Cycle for AI in E – Commerce at Daraz Sri Lanka, ............................................ 47
1.2.3 Role of the Project management and Project Team ..................................................................... 53
1.2.4 Work Breakdown Structure (WBS) ............................................................................................. 55
1.2.4.1 Work Breakdown Structure (WBS) for AI integration in E Commerce at Daraz Sri Lanka, ... 58
1.2.4.2 Project Milestones, (according to WBS)................................................................................... 64
1.3 The Cost Plan for AI integration in E – Commerce Optimization at Daraz Sri Lanka. .................. 67
1.3.1 ROI (Return on Investment) ........................................................................................................ 72
1.4 Time Plan ................................................................................................................................. 75
1.4.1 TimeLine for the Daraz E – commerce AI integration project TimeLine, .............................. 76
1.4.2 Examining the Time Plan............................................................................................................. 78
1.4.3 Gantt chart.................................................................................................................................... 80
1.5 Risk Analysis .................................................................................................................................. 83
1.5.1 Risk Analysis of Daraz’s AI implementation project .................................................................. 84
1.6 Quality Plan ............................................................................................................................. 89
1.6.1 The AI quality Control Framework of Daraz............................................................................... 90
1.6.2 Quality Plan for AI in E – Commerce at Daraz Sri Lanka........................................................... 92
1.6.3 Quality Control Strategy for Daraz Sri lanka’s AI integration .................................................... 95
1.6.4 Quality Management Tools that selected for Daraz’s AI implementation project..................... 103
1.7 Stakeholders .................................................................................................................................. 107
1.7.1 Key Stakeholders in AI for E commerce at Daraz sri lanka project .......................................... 108

15 Thanoj E127966
1.8 Resource Plan................................................................................................................................ 111
1.8.1 Resource plan for Daraz’s project .............................................................................................. 112
End of Task 1, ..................................................................................................................................... 115
Task 02, ............................................................................................................................................... 116
Literature Review................................................................................................................................ 116
Population of Daraz ............................................................................................................................ 125
Sampling Methods .............................................................................................................................. 127
Data types............................................................................................................................................ 135
Questionnaire ...................................................................................................................................... 158
Activity 3 ............................................................................................................................................ 170
References .......................................................................................................................................... 176

Table of Figures
Figure 1.Daraz....................................................................................................................................... 20
Figure 2.Sample E-Commerce Company Structure .............................................................................. 25
Figure 3.Daraz's Purpose ...................................................................................................................... 26
Figure 4.Daraze's Vision ....................................................................................................................... 28
Figure 5.How AI change the E-Commerce game ................................................................................. 29
Figure 6.Project lifecycle. ..................................................................................................................... 40
Figure 7.Phases of Project Lifecycle..................................................................................................... 41
Figure 8.Initiation phase ....................................................................................................................... 42
Figure 9.project planning phase ............................................................................................................ 43
Figure 10.Project execution phase. ....................................................................................................... 44
Figure 11.project Control phase ............................................................................................................ 45
Figure 12.Project close out phase. ........................................................................................................ 46
Figure 13.ROI Calculation .................................................................................................................... 73
Figure 14.Risk matrix ........................................................................................................................... 83
Figure 15. Fishbone diagram .............................................................................................................. 104
Figure 16.Types of Stakeholders ......................................................................................................... 107
Figure 17.Sample Resource Plan ........................................................................................................ 112
Figure 18. Daraz.lk ............................................................................................................................. 126
Figure 19.Simple random sampling .................................................................................................... 128
Figure 20.Stratified random sampling ................................................................................................ 129
Figure 21.Cluster sampling ................................................................................................................. 130
Figure 22.non-probabilistic sampling methods ................................................................................... 132

16 Thanoj E127966
Table of tables,
Table 1.Project Management Table ...................................................................................................... 40
Table 2.Roles and Responsibilities of Project management team ........................................................ 55
Table 3.Work Breakdown Structure ...................................................................................................... 57
Table 4.Cost plan ................................................................................................................................... 71
Table 5.Timeline of the project ............................................................................................................. 77
Table 6.Timeline .................................................................................................................................... 82
Table 7.Gantt chart ................................................................................................................................ 82
Table 8.Resource plan (Human resource) ........................................................................................... 112
Table 9.Resource plan (Technical resources)....................................................................................... 113
Table 10.Resource plan (Financial resources) ..................................................................................... 113
Table 11.Resource plan (Operational costs) ........................................................................................ 113
Table 12.Resource plan (Contingency and Miantenance) ................................................................... 114

17 Thanoj E127966
Activity 1,

1.1 Introduction to AI in E-Commerce,

1.1.1 How AI is Revolutionizing E-Commerce?

The greatest precious resources mankind has ever had, intelligence, is becoming a utility due
to artificial intelligence. Simply ask ChatGPT, Grok, or Gemini a question, and the language
models will respond after consuming a vast amount of human knowledge on the topic.
Only now are the consequences of this beginning to become apparent. My attention was drawn
to a comment from a PwC leader that made me consider how AI is affecting our worth and
usefulness as working people. In summary, the idea was that Gen-Z workers should be in the
workplace five days a week if they want to thrive in the AI era. That simply isn’t happening
now.
When intellect itself becomes a utility, what is the largest asset humans still possess? Without
a doubt, it’s human connection, a domain yet to be masteed by AL. This building interpersonal
relationships and may allude to the advantages of physical office locations and community-
focused activities in doing so.Idenfying leaders and research firms agree, identifying AL as a
critical focus area for driving business strategy and innovation. In 2024, the e-commerce sector
is at a critical juncture, largely because of its early adoption of AI technologies. The impact of
AI on e-commerce is expected to be monumental, offering unprecedented opportunities for
innovation and growth.
But the marketing environment has its own set of difficulties. This year, political
advertisements will cost close to $15 billion, a 31% increase over 2020 election cycle. At a
time when firms are attempting to market their way back to pre-pandemic engagement rates,
this will drive up the cost of digital advertising.
Futhermore, 70% of marketers worry that the efficacy of digital advertising will decline when
the third-party coocke is deprecated, which is expected to happen in !4 of this year. These
obstacles simply highlight how crucial it is to make a strategic change in order to survive, much
alone try to flourish.
It is obvious that 2024 is the year that marketers and brands need to concentrate on leveraging
their owned channels and ceating first-party datasets. Consumer may be reached through
channels like SMS and email, and first-parrty data enables businesses to establish a personal
connection. At a fraction of the price of sponsored retargeting, adding a powerful identity
resolution partner improves the capacity to opt in and reach more unidentified consumers
across multiple channels.
Don’t be fooled by today’s vendors who are merely adding Clippy-like AI enhancement to the
outside of their core products rather than gutting them from the inside out to rebuild with AI at
the center of their engine. The majority of marketing solution providers are promoting their
adoption of AI as revolutionary, but in reality, it’s like putting a new coat of paint on an old car

18 Thanoj E127966
-it looks shiny and new, but it doesn’t change its core function or value much at all. Remember
Microsoft’s clippy, the AI assistant that no one requested but was forced to incorporate into our
daily work streams? These vendors won’t be on the cutting edge of what is happening in 18
months.
The adoption of AI by E-Commerce companies is not cantered on these Clippy-like
productivity gains in applications designed for employees. With a large portion of the human
process eliminated, the main focus should be on business results, such as more revenue. This
indicates that AI is poised to upend established business paradigms in addition to causing
significant technological disruptions.
When the clod was introduced, software-as-a-service (SaaS) firms that offered yearly
subscriptions to software housed on the Cloud upended all of the established vendors who were
selling on premise software with perpetual licences. Simply said, it was more affordable, and
you could terminate your memebership if you weren’t happy with the service. The emergence
of cloud-native providers like salesforce practically tool off several on-premised software
heavyweights.
I think AI, which promises commercial results with little to no human engagement required for
many of today’s activities, will do the same. New vendors will appear with business models
centered on these AI-driven results rather than ones that rely on subscriptions for your staff to
work completion. For many of the tasks we currently perform manually , an application that is
solely AI driven will outperform any “tooling” designed to help marketers in their work.
However, in order to allow any given plan and achieve the intended result, that future depends
on AI learning from the appropriate data. This, in our opinion, is the foundation of
Wunderkind’s AI-first value proposition. In addition to storing the insights and context from
more than 2 trillion digital transactions, our identification network detects more than 9 billion
consumer devices and 1 billion opted-in customer profiles annually. When discussing the
results AI can provide, the data is what counts.
The business concept is straightforward; rather than a tool subscription, the goal is assured
revenue. The future of e-commerce and probably many other businesses lies in services that
use AI to drive an outcome. Additionally, marketers anticipate that the technique of creating
campaigns and consumer segmentation by hand and the concept of linear journey orchestration
will soon be obsolete. AI can accomplish this on a never-before-seen scale while eliminating
humans from the great bulk of the operation. Trillions of customer signals will be heard by
autonomous engines, which will deliver individualized, real-time experiences.
Businesses who invest in R&D to integrate AI into their core product and use their extensive
proprietary datasheets will be able to provide the results that brands require to succeed in the
new environment.
To guarantee that the human connection between the brand, solution providers, and their
consumers continues to thrive, all Gen-Z needs to do is return to the office.

19 Thanoj E127966
1.1.2 Introduction to Daraz,
Daraz Sri Lanka is the biggest as well as the leading e-commerce platform of the country which
provides to the customer a simple and easy to shopping environment. Daraz has been there
since its inception in 2018, serving as a catalyst for change to Sri Lanka’s online retail space
by offering a wide variety of products, helping local businesses and entrepreneurs thrive.
The company offers a marketplace enabling the buyers and the sellers to make online shopping
safe, efficient and fun. Daraz’s dedication towards bringing e-commerce to the masses, has
allowed it to offer security to all payments, reliable delivery to customers and innovative
shopping campaigns have driven countless millions of customers from all over the country to
daraz.
Daraz, backed by Alibaba Group, enjoys the cutting edge technology, logistics experience and
global e commerce best practice. Such was the company’s sustenance, they have been able to
offer DarazMall (selling authentic brand certified products), Daraz Express (its in house
logistics network) and Daraz Pay (a digital payment platform) to create a holistic, trustworthy
shopping environment.
Apart from retail, this year, Daraz helped add another leg to Sri Lanka’s digital economy by
helping SMEs go online as well. Daraz have employed it by way of seller education programs,
logistics support and financial support contributing employment and economic development to
the country.
In addition to traditional e-commerce, Daraz is known for its exciting mega sales events that
include the world’s leading e-commerce day, 11.11, 12.12, Black Friday and bring unmatched
discounts, exclusive deals, and interactive shopping experience to Sri Lankan consumers.
As Sri Lanka progresses further through digital revolution, Daraz remains as an affordable,
accessible, and innovative shopping solution provider. Daraz Sri Lanka, with a promise to
satisfy customers, digital transformation and empowering businesses, remains the leading
online retail destination in the country.

Figure 1.Daraz

20 Thanoj E127966
1.1.3 Daraz Sri lanka Company Structure

So as to ensure that Daraz Sri Lanka’s E-Commerce operations, shipping, and customer support
are effective, Daraz Sri Lanka maintains a formal hierarchical organizational structure. This
structure ensures that business process is on the track made by an organization in its strategic
plan and creates the ease of cooperation between different devesions. Daraz Sri lanka is
described in parts of the organizational structure as below,
1. Director Board,
Daraz Sri Lanka’s Board of Directors governs strategically and oversees Daraz Sri Lanka from
the top of the hierarchy. It is made up of senior executives, industry experts, and Alibaba Group,
Daraz’s parent company, representatives.
Rakhil Fernando – Managing Director and country head of Daraz Sri Lanka.

Kaushal Mendis – Chief Human Resources Officer at Daraz Pakistan and Sri lanka.

James Dong – CEO of Lazada Group, appointed as acting CEO of Daraz Group.

Bjarke Mikkelsen – Stepped down as CEO of Daraz Group.

Sajid Suhile – Director of Operations.

Veranja Gunawardana – Finance Lead Sri Lanka.

Mahela Liyanage – Regional Head of Legal.

2. CEO (Chief Executive Officer),

Daraz Sri Lanka’s CEO is answerable for the success and total management of the company.
The CEO reports directly to Board of Directors, who guarantee that the company is conducting
smoothly and operating in strict accordance with the corporate strategies, financial goals, and
industry trends in e-commerce field.
CEO – Rakhil Fernando
CEO’s responsibilities,
• Supervising the general growth and development of the business.

• Monoring the strategy’s implementation as well as the financial results.

• Encouraging innovation in digital transformation and e commerce.

• Serving as Daraz’s spokesperson in partnerships, public relations, and dealings with the
governments.

21 Thanoj E127966
 3. Team of Executive Management,
Begin the second section by introducing the senior level managers (Executive Management
Team) that handle specific part of Daraz Sri Lanka’s operations. Such professionals make sure
that the company continues to perform and bring a world class shopping experience to you.
COO (Chief Operating Officer),
COO is involved in working the day-to-day operation of the company and ensuring efficiency
especially in logistic, warehousing, customer service, and general operational process.
COO -Darshika Attanayake
Key duties,
• Overseeing supply chain and werehousing operations

• Overseeing Daraz Express, distribution network, and logistics.

• Enhancing customer satisfaction by increasing service quality.

CHRO (Chief Human Resource Officer),

Daraz’s CHRO manages HR’S functions, including recruitment, training, and assessing of top
talent that will lead to the company’s success.
CHRO - Kaushal Mendis
Key responsibilities,
• Supervising the hiring, training, and performance evaluation of staff.

• Preserving employee health and cultivating a good work environment.

• Preserving adherence to HR best practices and labor legislation.

CCO (Chief Commercial Officer),

Daraz Sri Lanka’s Chief Commercial Officer initiates the company’s commercial strategy
consisting of sales growth by sales and product offerings, vendor partnership to increase
revenue and market expansion.
CCO - Oshan Ranatunga
Keu responsibilities,
• Researching commercial strategies that will generate growth in revenues and expanding
the market.

• Maintaining product categories and vendor relationships by means of maximizing sales

and offerings.

22 Thanoj E127966
 4. Retail and E-Commerce activities,
In this regard, Daraz is a multi-category online marketplace backed with its e-commerce
division, which delivers a smooth shopping experience to its customers.
Team that manages the Category,
Category Head – Abdul Basith
• He oversees the different product categories (electronics, fashion, home appliances,
etc.).

• It works with sellers to keep quality at the right price.

Business Development and Seller Management,

Manager of Business development - Dhanushka Chathuranga
• It consists of its own team of recruiters and supporters who recruit and support 3rd party
sellers on Daraz platform.

• It trains and provides resources to sellers to perform better in your sales.

Customer Support and Service,

• Handles buyer complaints, inquiries and after sales services.

• Resolve disputes, refund policies, returns policies, handles.

5. Supply chain management and logistics (Daraz Express or DEX)

The delivery and logistics service operated by Daraz Sri Lanka is Daraz This ensures fast and
efficient deliveries through Express (DEX).
Director of Operations – Sajid suhile
Servig as the Senior Logistics Manager – Lahiru Wijethunga
Executive in Logistics – Daham Rathnayaka
Logistic executive for first Mile operations – Afham Nawfel
Warehouse and Fulfilment Centers,
• Retains and stores products for fulfilled by Daraz (FBD) vendors and DarazMall.

• Guarantees that clients receive their purchases on schedule.

23 Thanoj E127966
Last-Mile Delivery Network,
• Manages Daraz’s fleet to deliver products to client’s homes all around Sri lannka.

• Work with outside courier services to expand your reach.

6. Financial Services and Digital Payments (Daraz Pay)

Daraz Pay, an easy and secure digital payment solution is being introduced by Daraz Sri Lanka.
That division ensures that there are no problems in transactions and financial services for users.
Finance Lead for Sri Lanka – Veranja Gunawardana
Head of payments and partnerships – Dulika jayamanne
Key functions,
• Arranging cash-on-delivery mobile wallets, and credit/debit card payments.

• Putting in place fraud detection mechanisms to ensure asfe transactions.

• Collaborating with banls to offer discounts and EMI alternatives.

7. Marketing, Sales and Customer engagement

As a whole, Daraz heavily relies on digital marketing and data analytics to generate sales.
Head of Traffic Operations and Marketing – Firaz Marker
Head of Advertising – Atheeq kabeer
Senior marketing executive – Abdul Rahman
Key departments,
• Performance Marketing Team – Oversees Facebook, Google, and influencer programs.

• Affiliate & Partner Marketing – Collaborates with content producers and companies to
raise awareness.

• Live Commerce & Gamification – Uses interactive activities and live retail events to
engage shoppers.

24 Thanoj E127966
 8. Compliance, legal and Corporate Affairs
This department is to ensure that Daraz conforms to local laws, taxation policies and regulatory
instructions.
Regional Head of Legal - Mahela Liyanage
Head of Legal and Data protection officer – Piyumini jayawardana
Chief Corporate Affairs – Rakhil Fernando
Key functions,
• Overseeing contracts with partner and sellers.

• Making sure that consumer protection laws and e-commers regulations are followed.

• Overseeing data privacy laws and cybersecurity compliance.

Figure 2.Sample E-Commerce Company Structure

25 Thanoj E127966
1.1.4 Aim of the Project

The purpose of this study will be to analyze the organizational structure of Daraz to make sure
that it operates in an effective manner in e-commerce operations, for smooth logistics, and for
a healthy business expansion in Sri Lanka. This research will dissect the organizational
hierarchy of Daraz: where role of key management was, where the operational workflow was
& where strategy was taken.
Also, the investigation will be conducting testing on the effectiveness of the business model
that Daraz is adopting. This model involves the automation of inventory management,
detection of fraud, improvement of the supply chain, and the process of customizing customer
experiences by the AI (Artificial Intelligence). It will not only analyze how AI technologies
such as chatbot automation, data analysis, and machine learning support the communication of
a user, but will also consider the solutions such as order fulfillment and sales conversions.
In addition, the study will explore the challenges and advantages of AI application in e-
commerce Srilankan sector including data privacy, technologic infrastructure and customer
trust. In the research, the study will be conducted by taking the example of Daraz and looking
at how cutting edge e commerce platforms leverage AI to rev up operation pertaining to
businesses dealing in digital economy.
However, at last this study wants to give the Wholesale Sector such information about the
organization structure and the Artificial Intelligences strategies to improve the operational
efficiency and organizational position of the e-commerce, logistics, and tech livish sector in Sri
Lanka.
Daraz’s Purpose,

Figure 3.Daraz's Purpose

26 Thanoj E127966
1.1.5 Scope of the project

The main objective of this project is to study the implementation of Artificial Intelligence (AI)
for e-commerce operations at Daraz Sri Lanka, and their consequence on the implementation
of tariffs, customer experience and business efficiency. The goal of this study is to comprehend
how other artificial intelligence like powered technologies such as Machine Learning, Data
Analytics, Natural Language Processing are being used within industries to enhance
personalization, optimize inventory mom, to enhance fraud detection, and auto customer
support. Furthermore, it will look at the ways in which AI is facilitating the streamlining of
arrival of order, anticipating customer conduct, and bringing about business advancement in
Sri Lanka's advanced business exploit.
I take in the project as a case study for Daraz Sri Lanka, looking at how AI is used in different
functions of the platform. It will feature active AI based initiatives, e-commerce solution driven
with AI collaboration with technology firms and more importantly Daraz's contribution to
making Sri lanka better at advancing AI based e-commerce solutions. Furthermore, the study
looks at the positioning of the broader e-commerce sector in Sri Lanka, analyzing the key stake
holders in the e-commerce sector, including online retailers, logistics providers and such, to
assess their level of adoption of AI.
In addition, the research will also detail the challenges and challenges faced by Daraz and Sri
Lankan e-commerce sector in the adoption of AI including data privacy concerns, high
infrastructure costs, regulatory issues and ethical considerations. This will serve as a study of
how AI is going to change the e-commerce operation, and how it will help in improving
customer satisfaction and creating good shopping experience for customers.
This work evaluates the current state of AI within Sri Lanka’s e-commerce industry using both
primary and secondary research methods which include interviews with e-commerce
professionals as well as surveys sampled from the general population of e-commerce
professionals and garnered industry reports. It will offer actionable suggestions, which provide
recommendations to Daraz and other eCommerce companies in adopting AI, tackling
challenges, and utilising AI for long term business development. Insights learned will help
Daraz and similar e commerce platforms secure a seat at the forefront of AI powered digital
commerce in Sri Lanka.
Daraza’s Vision,

27 Thanoj E127966
 Figure 4.Daraze's Vision

1.1.6 Impact of AI concept on E-Commerce sector and Daraz

AI’s Impact on E-Commerce sector,

The e-commerce industry has been revolutionized by Artificial Intelligence (AI), which
improved a lot in customer experience, operational efficiency and business intelligence. These
include machine learning, natural language processing and predictive analytics and are using
AI to drive the online retail interaction with the customer, manage the supply chain as well as
optimize the marketing strategies.
Personalization was one of the most notable bearing marks of AI in e-commerce. At the same
time, AI algorithms based on huge amounts of customer information, including customer’s
browsing behavior, purchase history, and preferences, decide to promote products that exactly
match individual customers on what they should buy. As a result, conversions rates are higher
and customer satisfaction is improved. AI powered chatbots and virtual assistants also
transformed customer service options, allowing immediate response, assistance on order
tracking, addressing all the queries 24/7.
More than that, AI supports in detecting fraud and cybersecurity using unusual transaction
patterns and preventing fraudulent activities in real time. AI in supply chain management helps
in demand forecasting, dynamic pricing & inventory optimization, thereby eliminating excess
inventory and shortage of products, and also ensuring hassle free supply of products.
Additionally, automation driven by AI in logistics and warehouse management helps in
enhancing efficiency by making delivery routes more efficient and the warehouse operations.
Employing AI in the areas of advertising efforts, market trend analysis and development of data
driven marketing strategy helps businesses in coming up with their better, focused target group,
thus helps in increasing their ROI. Through the use of these technologies, voice commerce,
combining visual search and AR shopping with AI, and AI more generally, they increase
customer engagement and lower expenses, particularly online stores.

28 Thanoj E127966
AI’s Impact on Daraz Sri Lanka,
Daraz Sri Lanka is one of the leading e-commerce platforms in the country and actively using
AI to improve operations, enhance customer experience and promote business growth. In fact,
AI is quite involving in different roles of the entire platform, ranging from personalized
recommendations to automated customer support to operational efficiency.
Recommendation engines] that powered by AI offer a suggestion of products based on
customers’ behaviors. The feature will improve customer satisfaction, increase engagement and
sales. Furthermore, AI powered chatbots and virtual assistants handle a large amount of
customer query, giving an instant support and reduces the workload of human in customer
service.
AI means that Daraz can optimize warehouse operations, order fulfillment and delivery
processes in logistics and supply chain management. With AI, it can predict flowing of demand
trends and thus help in effective inventory management, so that above demand products are
always in stock and there’s less excess of inventory. Advance last mile delivery is also
influencing by AI, route optimization, shortening delivery times and reducing costs.
In addition, AI has added strength to the detection of fresh involving the platform and improved
payment security. With AI capabilities, it is able to detect and stop fraudulent transactions by
studying what buyers are doing and what they have done in the past; while also creating a
secure takeover as the buyer as well as the seller.
AI has also helped marketing and advertising on Daraz with automated ad placement, customer
segmentation, personalized promotions etc. By analyzing the customer data, AI will deliver the
ads targeted at them and it will increase engagement and enhances conversion rates.
Daraz’s growth would be increasingly influenced by AI in the future – visual search will
become more sophisticated, voice shopping will be more accessible and customers’ shopping
experiences will be enriched with AR. Daraz will keep its competitive edge, use AI to
streamline operations, optimize costs and improve customer experience in Sri Lankan e-
commerce industry.

Figure 5.How AI change the E-Commerce game

29 Thanoj E127966
1.1.7 Risk and Benefits
Risks,
• Privacy and Data Security issues,
The challenges with large volumes of customer data in an e commerce operation include risk
of data breaches and misuse. For these, Daraz needs to be held compliant with data protection
norms and has to adopt strong security protocols like encryption and secure payment gateways
to secure the data of its customers from unauthorized access.
• Expensive Implementation,
To achieve a personalized shopping experience from a sellers perspective, inventory
management point of views, and an advanced customer service AI technologies are invested
heavily by Daraz. But smaller businesses may find it difficult to afford the costs involved in
procuring, assimilating and running these advanced AI tools.
• Employee Opposition to Change,
In chosing new e-commerce solutions, employees at Daraz who are not familiar with AI can
be reluctant to embrace the new solutions mainly because of technical difficulty or fear of job
loss. But Daraz takes this transition smooth and smooth with training, and integrates AI tools
that assist employees and help them learns to the new technology and help operations to be
more efficient.
Benefits,
• Enhanced Ability to make decitions,
AI driven analytics are used by Daraz to process high amounts of data and it is used in
operational areas like pricing, customer interaction, and inventory management to enable the
taking of data driven decisions which improves Operational efficiency and customer
experience as a result.
• Increased Efficiency in Operations,
The repetitive line of task like processing orders, inventory manage and dealing with customers
can be automated as such that Daraz can save a lot of time and it reduces the human errors.
• Enhanced Client Contentment,
At Daraz, we use AI to enhance experience with every interaction – provide the right suggestion
based on the consumers’ preference, offer the right incentive to improve the engagement,
understand the consumer and self serve their needs.
• Saving on Expences,
AI is used by Daraz to predict demand, manage inventory and run targeted marketing
campaigns saving costs and increasing return on investments.

30 Thanoj E127966
1.1.8 Objective of the Project

The primary goal of this is to examine the real-world uses and constraints of AI in the E-
Commerce industry using Daraz Sri Lanka as a case study. Ai is rapidly changing the game of
Onilr shopping experience by increasing user interaction, streamlining corporate operational
efficiency.. Despite its many advantages, AI also has certain drawbacks, like data security risks,
high implementation costs ans ethical considerations.
The purpose of this study is to examine how artificial intelligence (AI) technologies like
Machine Learning (ML), Natural Language Processing (NLP), Predictive analytics, and
automated decision-making support the expansion of Daraz Sri Lanka and the larger E-
Commerce industry. The study will also point out possible ways to improve the use of AI and
the obstacles to its acceptance.
The objectives as follows,
1. To investigate the influence of AI in increasing customer side experience on Daraz
Platform,

2. To evaluate AI’s impact within the domain of supply chain and logistics management.

3. To evaluate AI’s performance in cybersecurity and fraud detection.

4. To investigate Ai powered marketing and sales optimization strategies.

5. To determine the obstacles and constraints to Daraz’s deployment of AI.

6. To offer suggestions for enhancing Daraz Sri lanka’s AI usage.

This research will look at how AI can change the nation’s E – Commerce sector, concentrating
on Daraz Sri lanka’s AI powered business practices. It seek to give a fundamental idea of how
AI iproves customer experience, supply chain management, fraud detection, and platform
marketing dtrategies. It will also look at the disadvantages and restrictions of using AI,
Inlcuding issues with data security, integration, moral dilemmas, and the requirement for
skilled AI experts. By examining these variables, the research seek to clarify how Daraz may
firther levrage AI to optimize business procedures and maintain competitiveness in the quickly
changing digital market.
The goal of this study is to positions Daraz Sri Lanka as a pioneer in AI powered E – Commerce
innovation in Sri lanka and a major force behind raising customer happiness, operational
effectiveness, and company expansion. To ensure that Daraz can leverage AI’s promise while
overcoming obstacles, practical suggestions for the ethical and sustainable integration of AI
will be given. This project will give information on how Daraz may overcome adoption hurdles
and leverage AI to enhance marketing tactics. Fraud detection, logistics, and customization.
The study’s ultimate objective is to provide Daraz sri lanka a clear path for using AI, establish
a benchmark that other E – Commerce companies in sri lanka must adhere to, and encourage
the expansion of AI driven e-commerce.

31 Thanoj E127966
1.1.9 Purpose of the Project

Using Daraz Sri Lanka as the case study, this research aims to point out and evaluate the useful
uses, advantages, and disadvantages of AI in the E – Commerce field. AI is still transforming
many different fields, and its incorporation into E - Commerce has enormous potential to
improve user experiences, streamline processes, and spur company expansion. The purpose of
this study is to assess how Daraz’s core functions, such as demand forecasting, chatbot-assisted
customer support, inventory management, and tailored recommendations are being affected by
artificial intelligence (AI).
The research also will evaluate how AI may boost consumer happiess, promote productivity,
and facilitate data-driven decision making in sri lanka’a E - Commerce industry by looking at
a number of AI driven developments. The study will also examine the operational, financial,
and technological obstacles to integrating ai into a local E- Commerce environment, as well as
how Daraz is overcoming these obstacles to maintenance its competitiveness in the expanding
online market.
1. To investigate the influence of AI in increasing customer side experience on Daraz
Platform,
Daraz benefits from AI powered recommendation engines, chatbots, automated customer
support systems, and so on, in order to create a better user experience for their customers. These
are tools which use a data-analysis technique to give precise recommendations together with
prompt assistance, sentiment analysis to boost product recommendations, virtual assistance,
and customer satisfaction as well as sales.
2. To evaluate AI’s impact within the domain of supply chain and logistics management.
There are three components to having an effective inventory management, demand forecasting
and delivery route optimization: These AI analytics are powered by AI to forecast cognize
about purchasing patterns, minimize time required in delivering to customers and minimize on
supply chain inefficiencies. This study will evaluate how Daraz Sri Lanka is leveraging the AI
tool for logistics planning, warehouse automation and stock management to improve
operational efficiency.
3. To evaluate AI’s performance in cybersecurity and fraud detection.
With online businesses, these frauds are possible such as fraudulent transactions, account take
over the cyber threats. Fraud Detection using AI: AI based fraud detection system analyzes the
pattern of transaction and keeps track of the suspicious activities and stops them before it
occurs. In this study we will look at how Daraz Sri Lanka delves into the use of AI based risk
assessment, payment security, user authentication to put their platform and then boost business
success and resilience in the competitive market for long run.

32 Thanoj E127966
 4. To investigate AI powered marketing and sales optimization strategies.
Automated advertising, automated marketing to customer segments and automation of the
pricing models are mainly revolutionized e commerce marketing by AI. Creating targeted
marketing campaigns, ad placement optimizing and predicting consumer behavior are what
Daraz Sri Lanka uses AI to do. It will research the relationship between AI driven insights that
can help increase sales conversion rates, reduce the cost of customer acquisition, and optimize
digital marketing, to result in a more strategic and efficient marketing, in accordance with the
consumer’s wishes and in accordance with the market and trends.
5. To determine the obstacles and constraints to Daraz’s deployment of AI.
The benefits of integrating AI in e-commerce are so great that nonetheless there are issues such
as data privacy, very high implementation costs, integration complexities of the system and
ethical dilemmas. In this study, we will analyze the barriers to Daraz Sri Lanka using AI fully
and will give insights into how this is hindering their growth and sustainability and will propose
solutions to resolve these issues. However, to make fullest use of the AI’s potential in the e-
commerce industry, the first step is to precisely understand these barriers.
6. To offer suggestions for enhancing Daraz Sri lanka’s AI usage.
According to the results obtained from the research, this study will propose strategies for
implementing AI in different areas of Daraz Sri Lanka’s operations. The best practices of
improving AI customer service, optimizing logistics and inventory systems, improving fraud
detection mechanisms, and refining AI-based marketing campaigns will be recommended.

Thus, this project ends with a cross analysis of how AI technologies were adopted in e
commerce in Sri Lanka with a special emphasis on Daraz which is one of the top online
shopping platforms. The analysis is intended to investigate how AI is being used presently in a
wide range of e-commerce situations, for example, personalization guides, inventory
management, customer assistance, and fraud cutoff.
The project will determine the advantages and disadvantages of integrating the AI, in relation
to how an AI can help Daraz in making customer experience better, making their operations
smoother and making sales performance higher. Additionally, strategic suggestions will be
provided as to how AI technologies can be additional embedded to enhance product offerings,
improve customer engagement, as well as enhance the supply chain.
Findings of this project will be useful for e-commerce professionals, AI developers and the
policymakers in Sri Lanka as they will give valuable knowledge of the role of AI in e-
commerce today and how it can be used to drive innovation, build customer satisfaction and
long term growth in e commerce industry.

33 Thanoj E127966
1.1.10 Significant Aim of this project,

The contributions of this research are: (a) research on the transformative role of Artificial
Intelligence (AI) in e-commerce, and; (b) an examination of how Artificial Intelligence (AI) is
being practically deployed in e-commerce at Daraz, one of Sri Lanka’s biggest e-commerce
platforms. The project aims to be a revolutionary project when using AI to do revolution around
e-commerce in Sri Lanka, increasing operational efficiency and customer experience and also
retaining cost effective. Through this research, we will examine the different aspects which AI
pertains to within the business model of Daraz, either in its positive opportunity or challenges.
• Improve the Customized shopping experience,
Owing to AI’s ability to analyse huge volumes of customer data, such as their purchase
behaviour and activities, it can make personalized product recommendations. The scope of this
research is to find out how AI can help enhance the user experience on Daraz and supply
content that fulfills customers and sales.
• Enhance the nventory management
AI algorithms help with predicting demand trends, that the demand is going to happen, and
inventory optimization whereby products are available when they are needed but not too much
stock is accumulated. In the research, an experimental approach will be used to investigate how
Daraz can use AI to optimize its operations by removing inventory management’s heavy
operational burden, improving product availability, and cutting down on operational costs.
• Enhance Customer Support Using Chatbots
Real time customer support can be given by these AI powered chat bots by answering the
queries and solving the issues immediately. The aim of this research is to understand how Daraz
can introduce AI based customer service solution for maximum engagement of customers and
minimum response time.
• Improve the identification and prevention of Fraud
AI comes in helpful to analyse transaction patterns and establish irregular behavior that can
spot and reduce from fraudulent activities. The research aims to learn how Daraz can utilize AI
in such a way that transaction is secured and customers, as well as the platform, are protected
from fraud.
• Resolve Data Privacy and Ethical issues
When the times come for AI, it becomes a test for how to decrease data security, customer
privacy as well as algorithm transparency. The objective of this project is to figure out what
ethical and legal challenges Daraz would get itself into during Sri Lankan context and propose
guidelines in comprehending AI responsibly to protect the data of users.
• Improve Logistics and the Supply Chain
Also, logistics can be optimized by prediction of delivery times, optimization of route planning,
and optimization of the warehouse operation. The research will check out the methodologies

34 Thanoj E127966
by which Daraz can utilize AI to reach a far more efficient supply chain, making shipment
times faster and also offering much better general service.
• Promote Marketing Initiatives
Using AI, customers data can be used to predict trends, and better marketing strategy
campaigns can be created. The aim of this research will be to learn how Daraz can adopt the
use of AI in its advertising activities so that it delivers the content which is relevant and would
be engaging to its audience.
• Develop E – Commerce Experts
AI automates routine tasks and helps you in making strategic decels and complex problem
solving. The purpose of this research will be to explore how AI tools can help Daraz’s staff to
increase their productivity and the capability of decision-making in the field of e-commerce.

1.2 Project Management Concept

This project has been set to assess the applicability of Artificial Intelligence (AI) in improving
e commerce process with a cut into Daraz Sri Lanka. During this study, data related to customer
interactions, sales transactions, inventory management, market trend will be collected and
analyzed to evaluate how AI can be used in revolutionizing online shopping experience. It will
look into how the AI can be useful for recommending products according to customer
preference, customer services, pricing strategies and logistics among others, which will then
boost customer satisfaction and grow the business.
This project aims to understand if and how AI can augment some processes such as e commerce
operations to increase the efficiency and effectiveness of product discover process, customer
engagement, and logistics process in Daraz for Sri Lanka. It also analyzes how AI can help
online retailers use more precisely individual predictions of consumer behaviour, personalize
shopping experiences, and with better supply chain management, among other use cases. Apart
from talking about how AI is limited in e-commerce (in terms of data integration failures,
algorithm failure concerns and privacy for customers), the paper will draw attention to the lack
of AI adoption in the e-commerce sector.
Project Management Aspect Details

Project Name Practical Applications and Limitations of AI in E - Commerce

Project Sponsor Board of Directors of Daraz Sri lanka

Project Manager Thanoj

Last Revision Date 2025.03.03

Project Approval Date 2025.01.12

35 Thanoj E127966
Description of the Scope • The pur[pose of this research is to point out real-wold uses and
intrinsic constraints off AI technology in Daraz Sri Lanka’s E –
commerce activities.

• Some points to focus on in this regard include deploying AI based

chatbots for improving customer services and recommending
personalized shopping experiences with AI based engines.

• In addition to solving data privacy and biases problems, this project

also tackles the ethical concerns while trying to integrate the AI into
Daraz Sri Lanka’s current e-commerce systems.

• The aim of this paper is to give practical recommendations about

how the use of AI technologies can be implemented in the e-
commerce industry of Sri Lanka by means of Daraz AI strategies
towards customer satisfaction and operational efficiency.

Project Deliverables • Ai Powered E – Commerce Tools: Describes the implementation

of an AI solutions that includes Demand forecasting, Fraud
detection, and Personalised recommendations using consumer data
analysis in context with the evaluation of an AI solutions
undertaken at Daraz Sri Lanka. This will highlight the prevalence
of AI tools to improve product offering and fraud reduction.

• E – Commerce Optimization: This is a comprehensive guide on

how AI can be used to enhance Daraz’s e-commerce operations
using sales trend forecasting strategies, effective customer
segmentation in order to increase engagement and revenue growth,
and so on..

• Ethical and Technical Limitations Report: Detailed discussion

about what obstacles Daraz must overcome to incorporate AI
technologies such as barriers of data security, algorithmic bias and
technical difficulties. Future solutions proposed in the report
include tuition, employment, insurance, licensing, and generally
more equitable compensation for work, as well as ethical AI
practices.

• Training Modules:It is a set of tailored training modules

specifically intended for use by Daraz employees and to equip
them to utilize AI tools effectively. As for these resources, they are
going to realistically look at both the practicalities of AI use and
the moralism’s of how to legally weaponize consumer data.

• Implementation Roadmap: A strategic implementation plan for the

gradual integration of AI technologies at Daraz Sri Lanka. This
roadmap will include making an allocation of resources, planning
key milestones and timelines for a smooth transition and a
successful AI adoption at the business level.

36 Thanoj E127966
Acceptance Criteria • Accelerated E – Commerce Operations: AT least 30% less effort is
spent on optimizing inventory management and customizing
consumer experiences thatn with conventional approaches.

• E – Commerce Efficiency: Evidence that AI driven solutions have

increased conversion rates, decreased cart abandonment, and
expedited customer support.

• Ethical Compliance: Respect with International norms for AI

transparency in E – Commerce operations as well as Sri lankan
data privacy regulations.

• Stakeholder Satisfaction: After implementing AI solutions, Daraz

Sri lanka’s management, staff, and customers should be at least
85% satisfied

Constraints • Budget Limitations: Limited money for AI tool acquisition,

personal traiing, and long-term support in E – Commerce
operations.

• Technical Integration: Integration of AI technologies with Daraz

Sri Lanka’s current operational procedures and E – Commers
system is challenging.

• Regulatory Compliance: Ensuring conformity to local data

protection regulations and international standards for E –
Commerce and AI use.

Assumptions • Availability of Expertise: The project will have the necessary E –

Commerce, data science, and AI skills.

• Infrastructure Readiness: The current IT infrastructure of Daraz Sri

Lanka can enable safe data processing and AI technologies for E –
Commerce activities.

• Stakeholder Engagement: Throughout the project, Daraz Sri

Lanka’s management, staff, and customers will continue to
actively participate and support it.

Performance Monitoring • KPIs: Improvements in important indicators including consumer

and Reporting conversion rates, the effectiveness of inventory management, and
the result of customized shopping experiences.

• Automated Reporting Systems: Ai genarated reports are used on a

regular basis to track E – Commers success and pinpoint areas
where consumer satisfaction and operational effectiveness need to
be improved.

• Stakeholder Updates: Stakeholders and the management of Daraz

Sri lanka are regulatarly updated on presentations and progress
reports.

Safety and Compliance • Data security and Encryption: Application of strong data
Frameworks encryption methods to safeguard private client and transaction
information in online business transactions.

37 Thanoj E127966
 • Access Control Measure: Limiting authorized personnel’s access
to AI technologies and client data in order to maintain
confidentiality and stop abuse in E – Commerce operations.

• Ethical AI standards: Following rules for objective AI systems and

making sure consumer data is handles fairly in E – Commerce
operations.

Cost Management Budget allocated for,

1. Acquiring AI techniques for E – Commerce.

2. Programs that teach employees AI skilled and methods.

3. AI systems should be updated and maintained continuously for

operational effectiveness.
• AI systems should receive regular maintenance and upgrades to
guarantee peak performance and ongoing development in E –
Commerce operations.

Time Management Duration:6 months.

Stages and Significant Events,

• Research Planning (Weeks 1 - 3) - Get final scope, objective and

key deliverables for AI integration in e-commerce.

• Data Collection and Analysis (Weeks 4 - 6) - Do primary and

secondary research online on customer behavior, market trends, as
well as the use of AI in e-commerce.

• AI tool Evaluation (Weeks 7 - 9) - Look for and evaluate the AI

platforms which can be automated, personalized to customers, and
aids in optimization of operations.

• Training Module Development (Weeks 10 - 11) - Make employees

understand how to make use of AI tools in e-commerce operations
by developing training materials for them.

• Implementation Roadmap (Weeks 12 - 14) - Write something of a

phased strategy for AI integration, make some allocations with
resources, and set some key milestones.

Quality Management • Data Validaition - To ensure that data fed to AI algorithms for
decisions in ecommerce operations are accurate, reliable.

• AI Testing - Therefore, to make sure that the AI technologies

worked—namely, they were capable of forecasting demand,
predicting client behavior, and offering tailored
recommendations—thorough testing of AI technologies was
needed.

• Compliance - Making sure that you follow the data protection

regulations and AI instance in the e-commerce industry so it will
be credible and trustworthy.

38 Thanoj E127966
Communication Plan • Internal Communication - It is vital to the ensure that project
participants work together in smooth manner through regular team
meetings, email updates and collaboration tools.

• Stakeholder Communication - To present and report on the

progress of important decision makers, operators and management
of Daraz Sri Lanka.

• External Communication - Thereby already been to share expertise

within the larger industry and tech community, publish insights or
even give presentations about the advancement of AI based e-
commerce.

Risk Management Identified Risks

1. Limited Data Availability - The challenge to collect enough high

quality consumer data to use as input for insights by AI.

2. Resistance to change among employees - They are fearful, in other

cases, that AI technologies will either make them redundant in
their jobs, or they feel unequipped to use these technologies.

3. Algorithmic biasis and data quality issue - AI predictions may be

biased or erroneous and therefore, it yields less than ideal choices.

Mitigation Tips

1. Establish data-sharing agreements with trusted partners –

Collaboarate with reliable data sources to raise AI model’s
accuracy.

2. Conduct staff training and engagement workshops - Outline to the

staff the advantages of using AI and give some practical training to
make the transition without any hitches.

3. Regular sudits of AI algorithms - You have to be attentive to and

improve on AI models regularly, to ensure impartial, equitable and
generally more favorable results.

Resource Management • Human Resources – AI specialists, data analysts, software

engineers, and E – Commerce experts.

• Technical Resources – Cloud computing infastructure,

recommendation engines driven by AI, cybersecurity solutions,
and data analytics tools.

• Other Resources – Funding for the usage of AI technologies,

databases on consumer bahavior, and physical and digital
infastructure to support AI-driven operations are all accessible.

Stakeholders • Daraz Sri lanka management and board members

• E – Commerce speciatists and data analysts

• AI ezperts and software developers

39 Thanoj E127966
 • Regulatory agencies in charge of data privacy and AI
implementation in Sri lanka.

• Partners in technology for the creation and use of tools related to

AI.
Table 1.Project Management Table

1.2.1 Project Lifecycle

The PMBOK Guide, Sixth edition defines a project life cycle as the sequence od stages that a
project goes through from inception to conclusion.
A collection of logically connected tasks that lead to the compleion of one or more deliverable
is calles a project phase. Along woth thrie tites, numbers, and durations, the stages may be
iterative, sequential, or overlap. The phases of the project management lifecuycle are
determined by the nature of the project, its intended purpose, and the management and control
needs of the organization or organizations involved.
Every step of a project has a start and finish date as well as a control point. Reevaluating the
project charter and business documents is the control point’s goal in light of the current
circumstances. The performance of the project is then compared to the project management
plan to determine if it should be modified, stopped, or continue as planned.

Figure 6.Project lifecycle.

40 Thanoj E127966
Every project has a beginning and an end, and each project has different tasks and unique
delivarables. This is why the comparison is made. The project life cycle may be impact by a
number of factors, including those inside the organization, the sector, the development
methodology.
The lifecycle provides the foundational framework for project management, regardless of the
specific task carried out. A typical project life cycle consists of five phases, regardless of the
size, complexity, and variations of each project.

Figure 7.Phases of Project Lifecycle

A project management life cycle’s function as a guide is one of its advantages. For the project
manager and project team, the phases simplify the procedure.
In this manner, they are always aware of what to do next anf how to take projects from start to
finish. There are five project phases in the project management life cycle.
1. Project Initiation Phase,
The beginning phase is the initial stage of the project management life cycle. This stage marks
the beginning of each endavor when we muct decide favorably on the goals we must
accomplish.
There are a few stages incolved. You must first determine the main issue that the initiative is
intended to addres. It would be ideal if you first identified the stakeholders and then the
project’s scope. Following that, you may create a statement of work and a business case.

41 Thanoj E127966
 Figure 8.Initiation phase

To decide whether to move on with the project, a business case is utilized. It contrasts the
project’s possible expenses and advantages. The statement of work, the other crucial document,
includes specifics on the project’s goals, delivarables, and other aspects of its scope.
Key activities,
• Define the project and determine its scope.

• Preliminary study of stakeholders.

• Preliminary evaluation of viability.

In order to communicate the aim to the other members of the project team, we must be
extreamly focused on it from the beginning of the project. It has been demonstrated that the top
acheivers and those least impact by negative stress are those who set challenging objectives for
themselves.

2. Project Planning phase,

Without a plan, you have no control and no clue whether you are doing alright. This phase of
the project is the plan that informs you where you should be in the first place. You need to
kknow how long it will take, how much it will cost, and what has to be done.
Project planning in project management is the process of dividing big tasks into smaller, easier-
to-manage pieces, which can result in a more realistic schedule and eliminate the risk of
“ground rush”, which is a team used in parachuting to describe the moment when you the
ground rushes to hit you when you are unprepared in the final stages of a jump.

42 Thanoj E127966
 Figure 9.project planning phase

• At this point, we lose our fundamental trining since all of the energy is concentrated in
the last seconds, casing fear and damaged limbs. It may occur when we need to adjust
the project plan’s prorities. Making a plan phases aer included in the planning process:
Determine the project schedule, then break it up into stages with certain activities that
must be completed throughout each phase.

• To receive the best rturn on investment, estimate the project’s budget and decide how
much to spend. Assemble the materials and begin assempling the squad.

To maintain the project’s quality and timeline, you should think about potential risks and
quality problems and begin planning to mitigate those risks. Create a risk assessment that
enables you to quantify and qualify any predictable problems in a project. It is recognized that
by begin proactive, we can better handle problems that arise in our social or professional lives.

3. Project Execution Phase,

The project plan is put into action during the execution phase. This entails implementing it as
quickly as feasible. When it comes to project execution, the well-known saying
“procrastination is the thief of time” is used.
Of cource, doing it with other is only feasible under certain situations. Working as a team is
necessary to accomplish goals like project management.

43 Thanoj E127966
People must learn to rely on the help of others by intentioanly delegating and sharing the
burden. Keep in mind that individuals who have wionderfull friends have more fascinating
lives in their private life.

Figure 10.Project execution phase.

During the execution phase, project managers should concentrate on a few things: tasks should
be assigned to the right team members and given the necessary guidance and explanation on
completing them. This also demonstrates your respect and trust for the team, which will foster
strong personal relationships and personal loyality, which will in turn encourage people to take
ownership of specific taksks and feel part of the digger picture.
Key components,
• Carrying out tasks and producing project deliverables.

• Cooperation and commincation among the team.

• Continuous evaluation of the project’s development.

Since all parties engaged in aproject should be informed of the progress achieved during its
execution, communication with team members, clients, and stkeholders is essential.

4. Project Control Phase,

Monitoring the procedures while they are carried out tis the main emphasis of this project
management life cycle phase, which begin with the prior execution phase. Project manager’s
daily efforts to maintain project work on schedule are usually referred to as control.

44 Thanoj E127966
 Figure 11.project Control phase

Key activities,
• Tracking project operations in real time.

• Issues are identified and resolved.

• Modfifications to project plan as needed.

The best way to guarantee development and advancement is to monitor and evaluate project
performance. A project manager should make sure that team members fulfil the deadlines and
quality standards for the work they are given.
Cost is another important factor to control. The project stays on course by budget. Additionally,
you should scan your QA checklists, expenditure records, and change management
documentation. This allows you to double-check your plans and monitor the flow of your
resources and efforts throughout the project life cycle. You are able to identify bottlenecks and
initiate crucial discussions on process enhancements for project.
Establishing observable benchmarks to assess our development in life or a project puts us in a
better position to make necessary corrections if things do not go as planned. Alternatively, if
we have accomplish a major goal ahead of time, we might reward ourselves.

5. Project close out phase,

The learning phase of a project is referred to as project closing. During this stage, we gather
and preserve information about our accomplishments and weaknesses. This gurantees that we
won’t keep going down the same foreseenable route to failure.

45 Thanoj E127966
We frequently fail to lean from our mistakes and end ourselves in unhelpful relationships or
circumstances, which can send us into a downward spiral of self-criticism.
You will be ready to use all you have leaned for your next project if you continue to remain on
course once the project is over. Make sure that every project component is finished and that
nothings is left unfinished.

Figure 12.Project close out phase.

Key tasks,
• Final project review and evaluation.

• Deliverables are formlly accepted.

• Documentation for Project termination.

The closure process entails terminating supplier contracts, releasing employees and equipment,
and giving the customer the deliverables and the owners the paperwork. Along with allocating
the leftover funds for next budgets, you should also provide report of important stakeholders.
Emphasizing the value of formla project management and introducing information that will
help the business in the future are the two main advantages of the closing phase for project
completion.

(Anon., 2024)

46 Thanoj E127966
 1.2.2 Project Life Cycle for AI in E – Commerce at Daraz Sri
Lanka,

2. Project Initiation Phase,

Objective
Also in this stage, Once the scope and feasibility of AI integration at Daraz are assessed, ensure
that it is aligned with the business objectives and available resources. Engage key stakeholders,
such as department heads, with whom you could get their support and commitment for success
of the project.
Key Activities,
• Feasibiility Study: Assess the capacity of AI to integrate with inventory management,
customer satisfaction and general operational efficiency at Daraz and then figure out if
it can make these works better so to say.

• Identifying Stakeholders: Work with Daraz’s leadership, IT, marketing and customer
service teams, as well as external AI experts and suppliers to ensure the implementation
of AI is aligned with business’ needs.

• Business Case Development: Focus on creating a business case of anticipated benefits

and advantages (automated customer support, personalized recommendations and
control of inventory) to promote the adoption of AI in Daraz.

• Improved Outcomes: Generate and showcase these AI capabilities to help show how
they can help power better outcomes for Daraz’s e-commerce platform using predictive
analytics, fraud detection and demonstrating that a chatbot can really help.

• Initial Budget Estimate: Without having to guess how much systems and features will
cost, provide an estimate of costs to carry out necessary infrastructure upgrades,
employee training programs, and AI software licenses to help Daraz integrate with the
AI.

• Risk Identification: The list of potential risks includes algorithmic bias, resistance to
change and data privacy concerns, and you need to consider prop up strategies for how
these risks can be mitigated during the implementation of AI at Daraz.

47 Thanoj E127966
Anticipated Results,
• The business case and anticipated return on investment (ROI) of a feasibility study.
Draw a detailed business case of how to adopt AI at Daraz, explaining how the lower numbers
of improvement to customer experience, inventory management, and efficiency will lead to
higher return on investment.
• Make a plan to include stakeholders.
Create a strategy to get the top people in Daraz such as its leadership, IT, marketing and
customer service teams involved and on the same page in order to integrate AI.
• High-level plan for putting AI into practice.
Identifying the key phases of AI adoption at Daraz, from the selection of technology,
integration of system together with employee training, to avoid any fast failures in the
execution of its AI implementation plan in high level.
• Initial risk assessment.
First, there should be an initial risk assessment of potential risks such as data privacy issues,
AI bias and system integration to address.

3. Project planning phase,

Objective,
Create a detailed and reliable approach to AI implementation at Daraz with the guide to the
allocation of resources, integration of technology and risk management. With such a plan in
place, we can utilize resources more efficiently and tackle the question of data security,
integration of systems and AI biases.
Key activities,
• Define project scope: Three major components of the project: selection of AI
technologies, preparing data, system testing, and deployment. The right choice of AI
powered tools for product recommendations, personalized advertising, and customer
service optimization can be used by Daraz as an example.

• Set up a Work Breakdown Structure: Assign responsibilities to cross-functional

teams. In whereby marketing team would concentrate on customer engagement
approaches, AI specialists would implement the AI powered systems, and IT
department would integrate them flawlessly to the existing systems of Daraz. There is
a key contributor to the team on each role to make sure that the project does not have
any bumps.

• Choose AI technologies: Choose among several AI technologies like recommendation

engines for personalized shopping experiences, automated customer service solutions
— AI driven chatbots — and demand forecasting algorithms that can aid the inventory
as well as the sales strategies at Daraz in Sri Lanka.
48 Thanoj E127966
 • Make a Timeline: Creating a Gantt chart that defines phases, deadlines, and
relationships is needed. Having this time frame in place will assist Daraz in making
smooth integration of AI; from picking the technologies to testing and deployment,
remembering the milestones.

• Risk Mitigation Plan: Plan for risks like algorithm failure, system integration
problems, data security and so on. Projects will be duplicated, AI tools updated
continuously to ward off failure, security will be robust, vulnerabilities will be regularly
tested to increase failsafe.

• Budget and Resource Allocation: Set aside budget for necessary investment in cloud
computing, AI expertise and similarly the cybersecurity infrastructure. To deliver a
successful AI solution using the Daraz platform, they have to make sure that they have
ready made resources available like, cloud infrastructure to scale up their AI solutions,
investment in skills training of the staff to be able to handle AI technologies.

• Compliance and Legal review: Make sure that all AI applications meet Sri Lankan e-
commerce and data protection laws. Regular legal reviews are necessary for Daraz to
integrate AI technologies ethically, and to accommodate only secure customer data per
the Sri Lanka’s Data Protection Act among others.

Anticipated Results,
• Detailed project plan with deadlines and completion dates.
The key stages of implementing AI at Daraz, with the selection, testing, deployment and
monitoring deadlines for them to be completed on time.
• Report on the AI technology selection.
The report that discusses the selected AI tools including the recommendation engine and
customer service platforms that will work with Daraz’s goals.
• Provide a report on the budget breakdown and resource allocation.
Breakdown of budget for implementation of AI which includes costs of cloud computing, AI
tools, cybersecurity and staff training.
• Risk reduction and management plan.
Plan dealing with risk of system failures and data security issues, including testing, monitoring
and protection of data.

49 Thanoj E127966
 4. Project Execution Phase
Objective
Seamlessly implement and integrate Daraz’s online store with AI technologies for better
customer experience, better inventory management and optimal business operations.
Key activities,
• AI Tool Integration: As AI driven tools such as chatbots, recommendation engines,
fraud detection systems etc, integrate the Daraz platform, the life of a customer is made
easier and the operations of the merchants transformed.

• Data Preparation: Given historical sales data, customer historical activity logs, and
inventory records, will need to prepare this data to feed the AI models for a better
product recommendations and demand forecasting.

• Collaboration with Teams: Teach Daraz employees how to apply the use of AI tools
to serve customers, promote the products and enhance sales using the AI technologies
across departments.

• Pilot testing: To test AI features out on Daraz’s platform, create a small user group and
test the use of the technologies out, bring in feedback and evaluate performance to see
if the technologies work well before full scale deployment.

• System Optimization: Constantly optimizing AI algorithms based on feedback from

the pilot testing, as well as to enhance product suggestions, inventory forecasts &
overall system accuracy at Daraz.
• Infastructure improvement: Increase processing capacity, cloud storage, and securit
Initial experiments with AI features on Daraz’s platform show promising results, with
improvements in product recommendations, customer service, and fraud prevention.
• y protocols to enable AI features.

Anticipated Results,
• Daraz’s online store has successfully integrated AI technology.
AI tools that Daraz has put on its online platform, recommendation engines, chatbots, fraud
detection systems, have all been successfully implemented on the Daraz platform to increase
the user experience and operational efficiency.
• Findings from early experiments showing AI performance.
Early tests of Daraz’s platform proved promising in terms of fraud protection, customer support
and product recommendations.

50 Thanoj E127966
 • Completed training and adoption programs for AI.
To smooth out the adoption and make the most of the power of AI through AI-powered tools
in customer support, sales, and marketing, Daraz has carried out training programs for
employees.
• Enhancements to the system based on feedback from pilot testing.
Daraz has based its system enhancements, based on the feedbacks from pilot testing, so as to
make product suggestion material, inventory forecast and end user experience more accurate.

5. Project Control Phase

Objective,
AI performance should be continuously tracked at Daraz for immediate action to tackle any
problem faced by it so that technology can provide good support to business goals and help in
e-commerce operations in general.
Key activities,
• Performance tracking: To measure the AI’s success in Daraz’s platform, we track the
key performance metrics (KPI) such as inventory turnover, customer satisfaction
ratings, and conversion rate.

• Data qualify Audits: Regular audits will need to be conducted to make sure Daraz’s
AI models use correct, unbiased data and well maintained high data quality for better
decision making and performance.

• Analyse feedbacks: Review customer feedback regarding AI driven interactions, chat

bot conversations or product suggestions on Daraz platform to improve user experience
and AI interactions on the Daraz platform.

• Refinement and Modifications: Rinse and repeat AI algorithms used for fraud
detection and product recommendations so that they become more accurate and make
the overall operational efficiency better at Daraz.

• Ckeck Compliance: Along these lines, Daraz carries out regular compliance testing to
make sure that AI systems that work at Daraz adhere to both national and international
e-commerce regulations by ensuring the safe and ethical use of customer data.

Anticipated Results,
• An analysis of the effects of AI is included in monthly performance reports.
Monthly reports also need to include a detailed analysis of AI’s impact on Daraz’s performance
with improvements in the improvements in the customer engagement, sales etc.

51 Thanoj E127966
 • Using real-time data to continuously improve AI models.
Use real time data from operations of Daraz to refine AI models such that they adjust as per the
changing customer preferences and market trends.
• Records for audits of compliance.
Keep track of records that Daraz AI systems maintain in compliance audits to verify that they
follow legal and regulatory standards in e-commerce and data protection.
• Feedback-driven advancements in AI
Continuously improving AI powered features, such as product recommendations and customer
support at Daraz, based on the feedback from the users as well as the stakeholders, implement
changes to the AI systems.

6. Project Closure phase

Objective,
You will continue to work on the performance of AI at Daraz to fix any issues as they come up
in order to continue to support the business and to get the most out of overall e-commerce
operations.
Key activities,
• Project performance Review: After considering the outcomes of implementation of
the AI project at Daraz, evaluate them against the initial goals and KPIs and verdict if
it was successful or not, find what needs to be improved.

• Documentation on lessons Learned: Write a document on the challenges faced, the

successes achieved and the points of improvement during the AI project for the others
in Daraz to absorb the experience so that valuable lessons can be used for future AI
initiatives.

• Final Report and presentation: Present a comprehensive final report covering the
project’s results, financial viability, and recommendations for future AI developments
at Daraz.

• Transfer to Operations Team: Make the IT and business teams at Daraz capable of
running and managing the AI systems so that it can be used in long run and integrated
into the business seamlessly.

• Plan for Post-Implementation Support: In addition, develop a plan for ongoing AI

support at Daraz based on step training, software upgrades, and maintenance for regular
systems to ensure efficiency and effectiveness of it.

52 Thanoj E127966
Anticipated Results,
• Final Report on the AI project evaluation.
To deliver a final evaluation report that summarizes how well AI worked at Daraz and it’s effect
on business operations, also customer experience.
• Knowledge transfer and operational handover.
Smooth the knowledge transfer of the AI development team to Daraz operations team for the
necessary skill and information for them to work with AI technologies.
• AI maintenance and optimization strategy
Daraz’ AI systems strategy should include how to maintain and improve performance,
accuracy, and customer satisfaction for the AI systems with ongoing maintenance and
optimization.
• Ideas for future developments in AI
Recommendations are stated based on what needs to be provided in terms of AI innovation and
enhancement at Daraz with regards to future trends and technologies which need to be adopted
to remain on top of the game playing for e commerce.

1.2.3 Role of the Project management and Project Team

An important role of project management is to successfully integrate AI into Daraz’s e-

commerce operations. It makes sure everything falls into place—everything budget wise,
coming in on time, and arriving at where the business needs to be. Being structured ahead of
time very clear of and early on the expectations, the challenges that will be in place and the
implementation process itself is a good thing.
In the planning phase, tools like Gantt Chart and Work Breakdown Structure (WBS) are used
to break the project into manageable stages, assign the resources and set the time as realistic as
you can. Among other things, some key considerations to take into account include the impact
of AI on data privacy, what employees can expect for new AI tools, and how the systems can
be smoothly integrated. Daraz's advantage in a well-organized project management approach
prevents delays, minimizes the risks, and ensures that the real value of AI being delivered by
the Project.
Tracking AI’s Performance
To evaluate the value of AI, Key Performance Indicators (KPIs) are tested like,
• Higher sales conversions from AI-powered recommendations

• Better stock management through predictive analytics

• Improved customer service with AI chatbots handling inquiries faster

53 Thanoj E127966
It is important for the project manager to bridge this gap as he is the one who aligns AI with
Daraz’s business goals. But regular monitoring and adjustments present a good way to
constantly fine tune AI systems by observing what real time data and customer feedback gives,
so that changes to the implementation remain effective.

Role Responsibilities

Project Lead
• Help the Daraz Sri Lanka’s AI team develop strategies through AI-driven
products to offer product recommendations, make the fraud detection
system stronger and make the logistics more efficient to have the best e
commerce experience.

• Working to bridge the gap between Daraz’s business and technical teams
in the case of the integration of AI, to uphold the goals of Daraz’s strategic
objectives and improve customer satisfaction and operational efficiency.

• In Daraz, you will oversee the budget and performance metrics of the
project and monitor the project progress towards certain key milestones to
make sure that AI bring about integration is done in a time bound and in
budget.

• Avoid scenarios of risks including the issues related to the security of data,
the reliability of the system, and challenges in AI at Daraz Sri Lanka by
identifying and solving them.

AI Engineers and
Data Scientists • Develop and enhance AI models for predictive analytics, automated
pricing, and personalized shopping experience.

• Examine customer data to increase the precision of demand projections an

recommendations.

• Address AI related biases and maintain system performance

improvements.

E commerce
operations and • Use AI driven insights in Daraz Sri Lanka to get more optimized inventory
Logistics Team management, better warehouse storage solutions and reduce delivery
delays to combine an overall better slimmer verified operation.

• Using AI technologies to automate supply chain processes at Daraz will

enable reduction of costs and accurate, timely order fulfillment to enhance
the customer experience.

• Use AI based logistics systems to track the delivery route, have better
planning and monitor the real time delivery statuses to provide smooth and
efficient logistics system at Daraz.

IT and Cybersecurity
Specialists • Oversee the smooth operation of the AI tools’ integration with the
current Daraz systems.

54 Thanoj E127966
 • To make Daraz' AI driven chatbots better, to make customers happy in
customer support and engagement with chatbots to make customer
shopping with chatbots seamless and responsive.

• Daraz implements AI powered recommendation systems to improve

search functionality, personalization of product suggestion, and giving a
better user experience.

Marketing and
Customer Experience • Use AI technology to study and optimize ad targeting, improve customer
Team segmentation and create a more personal advertisement for marketing
campaigns, thereby increasing engagement and increasing the sales.

• Using AI driven chatbots to continuously Upgrade daraz’s bot to be

upgraded for providing real time customer engagement and support to
customers while maintaining a smooth journey.

• Daraz can also implement AI powered recommendation system to make

product suggestions more relevant and better user experience for overall
search results.

Business Executives
and Decision Makers • All the AI related initiatives in Daraz must be in-sync with the long term
goals of the company and market expansion plans for sustainable growth.

• Monitor the performance in terms of revenue growth and customer

retention and areas for improvement of the implementation of AI at Daraz.

• Utilizes data driven insights to make strategic decisions for Daraz and
integrate AI technologies in a manner so as to maximize their ability in e-
commerce sector.

Table 2.Roles and Responsibilities of Project management team

1.2.4 Work Breakdown Structure (WBS)

What a Work Breakdown Structure is?

A work breakdown structure (WBS) is a project management system, that divides activities or
projects into smaller, easier-to-manage parts. It is a visual tool that simplifies the planning,
organizing, and progress tracking of the entire project. Each job is given a unique identification
by a WBS, which then arranges them in a hierarchical framework that illustrates the
connections between each task and the deliverables that are associated with it.
A WBS’s primary objectives are to enable efficient project planning and administration, give a
clear grasp of the project scope, and list all of the work that has to be done.

55 Thanoj E127966
Characteristics of a WBS,
You may create a work breakdown structure (WBS) in two different ways: deliverable-based,
which cocentrates on specific tasks or aspect of the project scope, and phase-based, which
devides projects according to timing.
Depending on the kind of project you’re working on, the size of your team, and your current
workflow, each WBS will have a somewhat different appearance. The majority of work
breakdown patterns have a number of similar traits, though.
• Hierarchical structure
The WBS breaks down project into smaller, easier-to-manage parts in a hierarchical manner.
The project’s primary deliverables are represented at the top of the hierarchy, while each
deliverable is broken down into smaller, easier-to-manage activities at the lower levels.
• Unique identifier
To facilitate tracking, monitoring, and control of the project work, each EBS component is
given a unique identification, such as a code or number.
• Defined scope
By listing every activity required to achieve the project’s goals, the WBS establishes the
project’s scope. This makes it possible for project managers to efficient plan, coordinate, and
oversee the project activities.
• Visual representation
For ease of comprehension and communication, the WBS is usually displayed as a graphical
representation, such as tree diagram.

(Anon., 2023)

56 Thanoj E127966
 Table 3.Work Breakdown Structure

WBS functions as a vital framework at Daraz Sri Lanka for implementing AI solutions by
specifying each project point which enables complete process management. Through its
method of dividing large processes into manageable units WBS enables complete clarity
combined with organizational structure and defined plans for AI deployment success.
The advantages of Work Breakdown Structure become evident in planning and execution
because it stops projects from expanding beyond initial plans thus avoiding time delays and
cost overruns. By using WBS the project team can set their primary focus on crucial work items
such as AI-enabled recommendation system integration and supply chain analytics
optimization with initial emphasis on automated customer experience improvement.
Project risk management benefits tremendously from WBS because its structured format helps
teams identify issues like data protection matters along with system connectivity hurdles and
employee reluctance to AI-based choices as well as training requirements. The team gains
better control over risk mitigation through defining the project into sequential phases.
In general, WBS is considered an additional advantage as it improves the internal
communication among stakeholders like IT teams, marketing specialists, and logistics
departments, aligning on the objectives, milestones and progress tracking. This is a very
structured approach in which project transparency is improved, data driven decision making
and real time performance monitoring is ensured.
In relation to AI at Daraz, WBS has designed it so that with every step of implementation of
AI—model development, algorithm optimization to compliance, coordination with
stakeholders and final deployment—it stays as well structured and executed as possible. It
makes the process of AI adoption smoother and easier, and consequently improves the user
experience, increases operational efficiency and boosts business delivery.

57 Thanoj E127966
1.2.4.1 Work Breakdown Structure (WBS) for AI integration in E
Commerce at Daraz Sri Lanka,

1. Project Initiation,

1.1 Define Project Scope,

• 1.1.1 At Daraz Sri Lanka, established objectives for AI driven e commerce

advancements to make personalization a reality, increase efficiency and grow e
commerce at the same time, optimise inventory and increase customer engagement.

• 1.1.2 When these are integrated into Daraz Sri Lanka, anticipated advantages are
improved logistics, shorter delivery time, better personalised product
recommendations, greater quality assistance by using AI, and better user experience,
which leads to better seller and operational efficiency..

• 1.1.3 Among the limitations of the project in Daraz Sri Lanka, financial barriers for
upgrades in infrastructure, constraints for integration of AI in present systems, its legal
restraints for the installation of AI in terms of data privacy and compliance with Sri
Lankan e-commerce regulations is situated.

1.2 Identify key stakeholders

• 1.2.1 The Daraz Sri Lanka AI integration project is led by the management team, the IT
department, the AI technology partners, logistics team and in front of the customer
service team.

• 1.2.2 For the AI adoption process in Daraz Sri Lanka, the Management Team is
responsible to ensure a strategy alignment and the ROI, while the IT Department takes
care of the integration and system reliability. AI Suppliers offer tools and services to
work with, that lean towards scaleability and customization. AI optimizes inventory and
delivery Logistics Team and optimizes the delivery process but the Customer Service
Team optimizes the process of delivering the products using the new developed
technologies.

• 1.2.3 Before being ready for communication with its stakeholder, Daraz Sri Lanka has
set up direct reporting channels, set normative updates on AI project progress and is in
line with business goals. This promotes transparency, builds trust, creates ease of
communication and information flow among all the parties involved in the development
of such a platform and key outcomes coming out of it.

58 Thanoj E127966
1.3 Develop project Charter

• 1.3.1 Key concerns of AI integration are addresses in the system compatibility, high
model accuracy and data privacy are given by Daraz Sri Lanka. It must involve
selecting the correct AI tools, registering them for tests rigorously, and while following
the local data protection laws to reduce the amount of risks.

• 1.3.2 As the AI integration project is very crucial for the company to implement, Daraz
Sri Lanka is seeking stakeholder approval to make sure this project will align with the
company’s business objectives and will obtain required resources and support to its
implementation.

2. Projract Planning,

2.1 Gathering Requirements,

• 2.1.1 Daraz Sri Lanka will evaluate in the project planning stage its existing systems
and data infrastructure for the presence or absence of gaps and plan to how AI
integration can be implemented effectively on the existing infrastructure in the project.

• 2.1.2 To accomplish AI’ integration for our users and developers in Sri Lanka, Daraz
Sri Lanka shall outline functional requirements like automation in the customer service,
demand forecasting, and inventory management, and non functional requirements such
as system scalability, data security, and performance efficiency.

2.2 AI technology selection,

• 2.2.1 As part of its operations, Daraz Sri Lanka is going to evaluate what AI can offer
for fraud detection, optimizing logistics, delivering personalized recommendations,
among others, but at the same time keep in mind how AI can be practical while being
good to customers.

• 2.2.2 Through a product selection process, Daraz Sri Lanka will examine how AI
solutions will adapt, if at all, to the existing IT infrastructure and to existing products
to avoid disruptions to operations.

• 2.2.3 The technologies that would comprise the AI stack for e-commerce for Daraz Sri
Lanka include machine learning based demand forecasting, fraud detection and
machine learning based personalized recommendations. It would also have Add AI
powered Chatbots for customer service and also use analytics tools to optimise
inventory management, pricing strategy and other customer segmentation.

59 Thanoj E127966
2.3 Project Scheduling,

• 2.3.1 The project schedule for Daraz Sri Lanka’s AI integration will have research,
testing, deployment and training phases. For 2–3 weeks, research will be done to see
market trends and the AI tools. Fraud detection and recommendations will go 4-6 weeks
of testing and 2-4 weeks of deployment. To make sure that employees are ready to
handle and make use of AI tools, a 2 week training period will be undertaken finally.
This timeline gives a chance for smooth smooth AI solution implementation.

• 2.3.2 The key deadlines of Courseworks 12 include research completion by Week 3, AI

testing completion by Week 9, deployment of AI onto the platform by Week 13, and
completion of employee training by Week 15. The benchmarks make sure that the
project is structured and timed so that it meets business objectives.

• 2.3.3 On Daraz Sri Lanka, each project stage will see proper resource allocation, backed
by data analysts in research, IT teams on testing and deployment and HR team working
on training to facilitate ease of adaptation. This helps in ensuring time project
completion as well as achieving goals.

2.4 Risk Analysis and Mitigation,

• 2.4.1 Possible risks at Daraz Sri Lanka would be AI Model Biases, Cyber Security risks
and Regulatory Compliance challenges. They will be pinpointed early on, and we will
address the risks before integrating AI with no unexpected issues..

• 2.4.2 To mitigate against risks, Daraz Sri Lanka will develop mitigation methods that
involve regular compliance audits and continuous AI model improvement; all of which
will be in line with regulations and with the performance increasing over time.

2.5 Budget allocation,

• 2.5.1 AI training cost, software cost, hardware cost and operational changes will be
decided upon by Daraz Sri Lanka to ensure that the budget matches the project’s scope
and objective and budget is not wasted.

• 2.5.2 Therefore, as part of the AI integration process, Daraz Sri Lanka will allocate fund
in the budget for any unexpected expenditure that may arise, addressing any unforeseen
situations or changed.

60 Thanoj E127966
3. Project Execution,

3.1 AI model development,

• 3.1.1 For instance, Daraz Sri Lanka will train AI systems with regards to Transaction
history, logistical data and customer behaviour to improve operations, product
recommendation improvement and inventory management.

• 3.1.2 To improve dynamic pricing strategies with robust machine learning models, as
well as more precise product recommendations, so more customer interactions and sales
takes place in the Daraz Sri Lanka website.

• 3.1.3 In order to make sure that those AI models are reliable and accurate, Daraz Sri
Lanka will conduct extensive testing before deploying them at scale.

3.2 System Integration,

• 3.2.1 To enhance operational efficiency and customer experience, Daraz Sri Lanka
will integrate AI models with its existing platforms namely the website, mobile app
and logistical software.

• 3.2.2 Make sure the data seamlessly gets transferred from Daraz’s orders management
system to AI systems so that all the processes of order processing and inventory
management remain accurate and efficient.

• 3.2.3 Integrate AI systems with Daraz and test that it works on all the platforms Daraz
has in existence to keep them functioning on the others.

3.3 Deployment and Testing,

• 3.3.1 Set up server side environment to deploy AI powered features on Daraz platform
for testing and getting real time feedback before full scale roll out.

• 3.3.2 Implementing A/B testing of Daraz to see the differences in sales performance
and customer satisfaction between the AI enabled and AI non enabled versions of the
platform using its platform.

• 3.3.3 Perform task to optimize the AI based features in Daraz's platform by finding
performance issues and refinement of systems based on test result to run it smoothly
from the beginning for large scale launch.

61 Thanoj E127966
3.4 Staff Training and Documentation

• 3.4.1 Perform task to optimize the AI based features in Daraz's platform by finding
performance issues and refinement of systems based on test result to run it smoothly
from the beginning for large scale launch.

• 3.4.2 Help Daraz's technical team to get in detailed documentation on supported

ongoing maintenance and troubleshooting AI system for operating without any issues
and resolving any issues that may occur quickly.

• 3.4.3 Getting feedback from Daraz’s employees so that we can identify what to
improve with AI-powered tools and make them workable and sounding with operation
objectives to boost the whole efficiency.

4. Project Monitoring and Control,

4.1 Performance Monitoring,

• 4.1.1 To measure the effectiveness of the AI technologies over at Daraz, we would use
the key performance indicators (KPIs), such as the sales growth, customer satisfaction
and operational efficiency.

• 4.1.2 This helps Daraz to continuously enhance AI projections by analyzing customer

behavior data and keeping if in line with customer preferences to better fine tune
recommendations, predict trends more accurately, and aligns with their customer
preferences better.

• 4.1.3 Improve the AI model to eliminate any performance disparity, with the model
achieving the best accuracy and efficiency in predictions, product recommendation as
well as customer engagement.

4.2 Quality Assurance and Compliance

• 4.2.1 Regularly run AI tests to see that the models are still fair, and unbiased, and
accurate, never without consistency but also reliability of them.

• 4.2.2 In reviewing AI systems periodically, to assist data security laws and regulations
to make certain customer details are protected and there is trust.

• 4.2.3 Findings from the audit will be used to adapt AI features and models to ensure
improved accuracy, to address problems, and ensure alignment of AI to business
objectives, and compliance standards.

62 Thanoj E127966
 4.3 Stakeholder Review and Feedback,

• 4.3.1 Updates, reports and presentations to regular stakeholders of progress,

maintaining transparency and line to project goals.

• 4.3.2 Consider what type of feedback from employees, partners, customers would be
important to think about prior to integrating AI to make sure it works for them so that
their experience remains positive.

• 4.3.3 Such continuous refinements in AI strategies should be made as per ongoing

feedback, analysis of performance, as well as emerging trends to optimize the
outcome and to be in line with Daraz’s ever changing business criteria.

5. Project Closure,

5.1 Final Evaluation and Report,

• 5.1.1 Want to evaluate the success and overall impact of practicing AI integration and
to do this you will have to measure key performance indicators (KPIs) such as sales
growth, customer satisfaction and overall operational efficiency and help you identify
areas for further improvement should there be.

• 5.1.2 Record the lessons learned from the AI project in order to help and advance
future AI initiatives, which should, in turn, lead to better decision, risk and
implementation of project in later developments.

5.2 Handover and Maintenance Plan,

• 5.2.1 Turning control of AI systems over to IT and data science teams within the
organization to maintain technical competence, technical abilities, to develop and
continue to develop AI models and technologies within the organization.

• 5.2.2 Implement a continuous monitoring, support and updates system on AI models

to keep them up to date with best in business practices, accuracy and evolve when
new technology evolves.

• 5.2.3 The AI systems would be scalable and sustainable to future developments and
growth of the business and the technology.

The success of AI in e-commerce at Daraz Sri Lanka is the outcome of its local retail turn to
the digitalization. This project is designed to provide a total approach based on planning, risk
management, system integration, quality assurance, and deployment for the purpose of

63 Thanoj E127966
increasing efficiency and improving the entire customer experience, all supported by a Work
Breakdown Structure (WBS).
Daraz stands strong as a strong advocate of AI enabled innovations such as automated customer
care, demand prediction and fraud prevention that help to make processes more efficient, be
highly engaging with customers and maximise on revenue potential. This platform is using
machine learning, AI powered chatbots, and lots of data analytics to predict consumer
behaviour, optimally manage their inventory, and apply dynamic pricing to last minute trends
in the market.
Furthermore, the associated project oriented roadmap of the Work Breakdown Structure of
Daraz’s AI integration helps the vision of post implementation system upgrades and the
continuous continuous maintenance of AI capabilities in the platform. The structure in this
approach addresses data security risks, manages AI biases and ensures that employees fully
adapt to and use the AI tools introduced in the system.
Somehow Daraz Sri Lanka’s AI initiative has been a success, a benchmark for future digital
commerce innovations and unique solutions for retail customers, a huge achievement. And with
DA basis the best in class AI excellence at Daraz, coupled with the continued stakeholder
collaboration, Daraz is in a position to become profitable and sustainably profitable, setting a
new standard for shopping online in Sri Lanka.

1.2.4.2 Project Milestones, (according to WBS)

1. Milestones of Project Initiation,

Project Scope Completion,

This finalisation of Scope of Implementation is indicated as Project Scope Completion when
we are defining Daraz Sri Lanka’s AI project objectives, desired outcomes and scope of
implementation. It explains how AI will enable better use of customer experience, logistics and
business processes. Also included as outline in the scope document are limitations in which
data availability, regulatory constraints, and system compatibility are taken into consideration.
Stakeholder Acceptance,
All the key stakeholders must confirm the project charter at this milestone; it includes Daraz
Sri Lanka’s management team, IT specialists, AI advisors as well as the heads of its operations.
This would give us a cherry on top, on the alignment of project goals, responsibilities and
expectations. This is also accompanied to keep stakeholders informed from the first inception
of the project through to its complete lifecycle.

64 Thanoj E127966
 2. Milestones of Project Planning,

Completion of requirement collection,

Before an implementation, all the functional and non functional requirements of AI
implementation are gathered in an intrusive interview. This includes current operations, user
needs, technical feasibility and compliance of Sri Lankan e-commerce regulations. This
collected information makes sure that the products made to fit Daraz Sri Lanka’s business
environment.
Completed Technology selection,
The suitable AI technologies, tools and machine learning models are selected at this stage. The
consideration of compatibility with Daraz’s existing IT infrastructure, being scalable, and being
cost effective are of the key consideration behind it. The company completes the selection of
the AI technology stack that they feel is the very best to meet the business requirements like
customer support, fraud detection, and personalized recommendations.
Approved Project Schedule,
Research, development, testing and deployment deadline is established through a detailed
timeline and resource allocations. There are milestones and checkpoints to track progress and
to keep clear as one progresses.
Prepared Risk Management Plan,
AI model accuracy is threatened by data security threats and other risks; the solutions are
provided. All potential delays (including the failure of systems), and backup plans and
cybersecurity measures are formulated to contain them.
Approving the Budget,
A budget for the whole package including software, hardware, people, and their training is
finalized. The finance team of Daraz Sri Lanka reviews and approves the cost estimates of AI
tools, cloud computing, infrastructure upgrade, and contingency expense.

3. Milestones of Project Execution,

System architecture Design,

Our architecture of the AI system is represented by ‘how data is flowing’ from AI models, from
databases and to Daraz’s backend systems. This system blueprint also guarantees that AI
solutions will be seamlessly incorporated into e-commerce operations without disrupting usual
operating functions such as customer service, real time order tracking, analytical predictive
capabilities, and so on.
Development of an AI model Completed,
Based on historical data from Daraz’s platform, such as customer behavior, transaction trends
and logistics efficiency, AI models are trained to understand the points of improvement using

65 Thanoj E127966
AI models. Data preprocessing, choice of the algorithm, model optimization for good
predictions is all part of this process. Thus, functional AI models suited for the test are prepared.
Finalization of System integration,
Once the AI models are built, they are integrated into Daraz Sri Lanka’s existing platforms
including website, mobile app and the backend management system. These days, AI tools are
linked to the databases, operational workflows and properly linked to have proper functioning
conditions at the milestone of this.
Initial Implementation of the Prototype,
A prototype version of the AI powered system is deployed with well as used internally as select
users such as customer service representatives, logistics staff, and IT administrators. It helps in
detecting the bugs and areas for the usability improvements. Feedback gathered during this
phase is subjected to further refinements.
Completed training and Documentation,
Daraz Sri Lanka provides training to its employees and key stakeholders around use of AI
powered tools, which are completely on the ground before the whole deployment. On the level
of documentation, user manual, troubleshooting guides, etc. are provided to facilitate seamless
integration and adoption of AI system on the whole teams.

4. Milestones of Project Control,

Observation and Assessment of Performance,

This is also a milestone that guarantees Daraz Sri Lanka is still monitoring AI performance in
line with predetermined KPIs. Dashboards that focus on the measurement of the dashboards
consists of AI accuracy, response time and impact on customer satisfaction. These insights can
then be put to use in order to change the required settings to improve the AI systems for
efficiency and users experience.
Testing and Quality Assurance Completed,
Functional, performance and security testing are carried out to the AI system to ensure
reliability and meet Daraz Sri Lanka’s business needs. The system is cleaned thoroughly before
going live in order to rid them of any inconsistencies or errors and the system operates
smoothly.
Approved User Acceptance Testing (UAT),
Before going live, the AI system goes through real Daraz users, including customers,
employees, and teams even from the operational department. The system is made better and
better by paying attention to the user feedback to optimize performance and usesesibility, in
which the system meets the needs and expectations of all the stakeholders.

66 Thanoj E127966
 5. Milestones of Project Closure,
Final Evaluation and Acceptance,
The project is reviewed in a form that summarizes achievements, challenges and lessons learnt.
To approve with stakeholders, performance reports, cost analysis and final system evaluations
are presented.
Completely Deployed AI system,
The AI system is launched officially on Daraz Sri Lanka’s platform after successful testing and
refinement. The users get access to AI enabled features like, automatic chatbots, personalized
recommendations, fraud detection systems, etc.
Project Transfer,
Basically, it is passed on officially to Daraz Sri Lanka’s IT team to maintain AI system, which
makes the project end. Long term management is handed over all documentation, training
materials and system configurations.

1.3 The Cost Plan for AI integration in E – Commerce Optimization at

Daraz Sri Lanka.

E commerce automation driven by AI enhances consumer satisfaction, boosts output, and

streamlines corporate processes. A through cost study of incorporating AI technology into
Daraz’s operations may be seen below.

1. Tangible costs
These are measurable costs associated with purchasing AI technology and associated
infrastructure.
• License for AI software – AI powered recommendation engines, fraud detection tools,
and chatbot automation tools.

▪ Estimated Price – LKR 3,000,000 (USD 10150)

• Data Storage and Cloud infrastructure – Coordinating order histories, client

information, and datasets for AI training.

▪ Estimated Price – LKR 1500000 (USD 5075)

• System and Hardware Upgrades – Computer power and High-performance servers for
AI processing.

▪ Estimated Price – LKR 3000000 (USD 10150)

67 Thanoj E127966
 • Marketing Automation Tools – AI powered ad targeting (Facebook, Google, YouTube),
email automation, and campaign enhancements.

▪ Estimated Price – LKR 4500000 (USD 15225)

• Tools for Administration and Support – Document processing, logistical tracking, and
workflow automation.

▪ Estimated Price – LKR 1500000 (USD 5075)

2. Intangible costs
These are intangible costs that impact a company’s ability to grow, its reputation, and its usage
of AI.
• Employee Education and Adjustment – Employee education on recommendation
systems, chabots, and AI driven logistics.

▪ Estimated Price – LKR 900000 (USD 3045)

• Customer satisfaction trust – Potential risks include incorrect recommedations or

unsatisfactory customer service interactions in the event that AI misinterprets user
behavior.

▪ Estimated Price – LKR 600000 (USD 2030)

• AI integration Downtime - Lost sales might arise from brief service interruptions during
AI deployment.

▪ Estimated Price – LKR 15000000 (USD 50750)

• Controlling change – Altering operating processes and ensuring AI integration goes

smoothly.

▪ Estimated Price – LKR 7000000 (USD 24000)

3. Direct costs
Expenses that directly affect Daraz’s AI deployment and implementation.
• AI development and Personalization – Employing AI developers to improve fraud
detection algorithms, recommendation engines, and product search.

▪ Estimated Price – LKR 2000000 (USD 20000)

68 Thanoj E127966
 • Employee salaries - Compensation for technical experts, data scientists, and AI
engineers.

▪ Estimated Price – LKR 15000000 (USD 50000)

• Consulting and Outside knowledge – Using AI specialists to develop plans and evaluate
results.

▪ Estimated Price – LKR 3000000 (USD 10100)

• Collecting and Analysing information – Analysing customer behavior, search trands,

and previous purchases to improve AI.

▪ Estimated Price – LKR 1500000 (USD 5000)

4. Indirect costs
Ancillary costs necessary t support AI infastructure and project execution.
• Costs of Operational Utilities – Maintenance costs, internet, and electricity.

▪ Estimated Price – LKR 900000 (USD 3000)

• Space for Facilitates and Servers – TO train AI models, consider expanding data centers
or renting cloud storage.

▪ Estimated Price – LKR 1500000 (USD 5000)

• Compliance and Security – Making sure AI conforms with data protection and
cybersecurity regulations.

▪ Estimated Price – LKR 1200000 (USD 4500)

• Overheads in administration – Extra charges for staff assistance, general operating

expenses, and the office.

▪ Estimated Price – LKR 2000000 (USD 7000)

69 Thanoj E127966
 5. Suck costs
Earlier financial commitments that are still non-recoverable.
• Costs of developing an outdated system – Previous expenditures on logistics
automation and antiquated recommendation algorithms.

▪ Estimated cost – LKR 4500000 (USD 15000)

• AI pilot projects – Early AI experiments that were assessed but never put into practice.

▪ Estimated cost – LKR 3500000 (USD 12000)

• Instruction on Antiquated systems – Expenses related to staff training for antiquated E

– Commerce automation systems.

▪ Estimated cost – LKR 1600000 (USD 5413)

Phase Category Estimated Estimated

Costs (LKR) Costs (USD)
Tangible Costs AI software license 3,000,000 10,150
Data storage and cloud infastructure 1,500,000. 5,075

System and Hardware upgrades 3,000,000 10,150

Marketing Automation Tools 4,500,000 15,225

Tools for administration and support 1,500,000 5,075

Subtotal 13,500,000 45,675

Intangible Employee Education and 900,000 3,045

costs Adjustments

Customer Satisfaction and Trust 600,000 2,030

AI integration Downtime 15,000,000 50,750

Controlling change 7,000,000 24,000

70 Thanoj E127966
 Subtotal 23,500,000 79,825

Direct costs AI development and personalization 2,000,000 20,000

Employee Salaries 15,000,000 50,000

Consulting and outside knowledge 3,000,000 10,100

Collecting and Analysing 1,500,000 5,000

information
Subtotal 21,500,000 85,100

Indirect costs Operational Utility costs 900,000 3,000

Facility and server space 1,500,000 5,000

Compliance and security 1,200,000 4,500

Administrative overheads 2,000,000 7,000

Subtotal 5,600,000 19,500

Sunk costs Outdated system development 4,500,000 15,000

AI pilot projects 3,500,000 12,000

Instruction on Antiquated systems 1,600,000 5,400

Subtotal 9,600,000 32,400

Total 73,700,000 262,500

Estimated (LKR) ($)
Budget
Table 4.Cost plan

Total Financial benefits of the project,

Benefit Category Estimated Financial Benefit (LKR)

Faster Order processing and Fulfilment 18,000,000

Improved AI powered recommendations 10,000,000
Operational Efficiency Gains 12,500,000
Cost Savings from AI driven automation 15,000,000

71 Thanoj E127966
 Enhanced Customer experience and 7,000,000
retention
Total Financial Benefit LKR 62,500,000

Net Financial Benefits,

Description Amount (LKR)
Total Financial Benefits LKR 62,500,000
Total Project Budget LKR 73,700,000
Net Financial Benefit LKR -11,200,000

1.3.1 ROI (Return on Investment)

Return on Investment, or ROI, is a static used to assess how well marketing expenditures
produce income. It’s basic idea for all marketers who want to show the financial returns on
marketing expenditures in order to defend them. ROI measurement is a strategic tool for
maximizing the marketing mix and resource allocation in addition to being a financial indicator
of success. By identifying which methods produce the highest returns, it helps marketeers focus
their future marketing expenditures on the most efficient channels and techniques. Since a
company’s bottom line and position in the market may be greatly impacted by the effective use
of its marketing budget. It is important to comprehend and improve your investment.

Key concepts and Formulas,

1. ROI Formula – (Net profit – Marketing investment) / Marketing investment * 100 is
the fundamental formula for calculating ROI. The return on investment is shown as a
percentage via this computation.

2. The whole cost of obtaining a new client, including all marketing and sales
expenditures, is measured by the client Acquisition Cost (CAC). Total Marketing and
sales Expenses / Number of Ne customers Acquired is the calculation.

3. Client Lifetime Value (CLV) – An evaluation of the overall value a firm will gain from
its whole relationship with a client. Calculating CLV helps determine how much a
business can afford to spend on acquiring consumers.

4. Conversion rates – This statistic shows the proportion of users who complete a desired
activity, which aids marketers in determining how well their efforts influence
consumers to go through the sales funnel.
(Anon., 2025)

72 Thanoj E127966
Artificial Intelligence (AI) has brought a great change in the e-commerce market, and in doing
so, Daraz also did an analysis to judge Return on Investment (ROI) of implementation of
Artificial Intelligence (AI). Stakeholders can trace the impact of AI’s use for profitability
growth and cost minimization in the long term through the use of AI driven automation and
personalization. This will help Daraz to expand its business organically and for investments in
future technologies and to keep focusing on continuous innovation in competitive e-commerce
marketplace.
To streamline its operation and improve both customer experience and conversion rate, Daraz
uses AI to optimize product recommendations, fraud detection, and automatically provides
customer support. However, there are two important points: substantial up front investments
are needed, notably, in software licensing, hardware investment, employee training, data
infrastructure, system integration. They are essential initial costs for having a robust AI driven
e-commerce platform setup.
If Daraz’s AI project delivers a positive ROI, apart from successfully recovering the
investment, the company will get added business value from increased sales, improved
customer retention, as well as reduced operational cost. High initial costs are typical in the
short term with a negative ROI, but the long term positive ROI obtained through cost savings,
improved efficiency and improved competitiveness of the market, justify the investment. This
ROI assessment will help Daraz’s managing team formulate the subsequent area of AI
investment after they map to the overall strategic endeavor of the firm.
ROI Calculation,

Figure 13.ROI Calculation

ROI = [(62,500,000 - 73,700,000) / 73,700,000] × 100

ROI = (-11,200,000 / 73,700,000) × 100
ROI = -15.19%

73 Thanoj E127966
For the first year, the ROI of making Ai integration at Daraz is -15.19% mainly due to the high
implementation costs, such as the cost of creating the Ai system, educate the staff for that, and
upgrading the infrastructure. However, this is a once-off expenditure and all further costs are
in terms of maintenance only and hence, a continuous profitability.

Due to the specific long-term gains in terms of automated processes, AI personalization in

products, operation efficiency, Daraz is expected to achieve a substantial return on investment
in coming years.
Financial Impact and AI integration at Daraz,
1. Initial cost difficulties,

• Large-Scale Investment - Tight spending on AI automation, machine learning tools, the

system integration as well as workforce training will consume all the first year. An
initial negative ROI is ensured by these expenses.

• Typical of AI Adoption - Like in other verticals, where AI projects are concerned it

often takes time to get to returns since optimization and wide scale adoption happens
over time.

2. Long Term Profitability and Value,

• Improved Commercial Activities - In turn, your customers will be interacting with AI

for orders, inventory management, delivery logistics which, in result, will be quicker
and cheaper.

• Growth Revenue and cost Reductions - These 2 tactics with AI powered

personalization, fraud detection, dynamic pricing strategy will help Daraz improve
sales, better customer retention, and better operational efficiency.

• Positive ROI projected - After the first year, the rest are mainly in terms of maintenance
to make it profitable in the long run with maximum financial benefits.

3. Efficiency and Market Domination,

• Effective Order Processing - The automation will fuel the orders without compromising
quality or turnaround time, by reducing the fulfillment time and minimizing the errors.

• Groeth and Scalability - Better Forecasts, Better Supply Chain, Less Returns Making
The Long Term Financials Sustainable.

74 Thanoj E127966
 4. Gaining a competitive edge and building brand trust,

• Pioneering innovation in E – Commerce - The integration of AI will place Daraz as one

of the frontrunners in the region when it comes to numbers of sellers, customers and
investors.
• Customer Trust and Loyalty - Recommendation systems will be improved, chatbot will
be supported, fraud prevention will be done, which is good for user experience, those
will improve the brand reputation and market dominance.

1.4 Time Plan

Analyzing and creating a timetable and schedule for project completion are key components of
project time management plan. Project deadlines that are either overestimated, as well as
unforeseen obstacles, are mitigated by formalized time management procedures.
Plans for time management establish which tasks should be modified as well as how resources
should be distributed and managed during the project.
Project managers have different definitions of project time management, and different project
management approaches have different techniques. As needs change, for example, an agile
marketing approach may demand work that may be completed in brief sprints. On the other
side, managers who employ kanban techniques may schedule time in more liberal chunks,
facilitating a more seamless worlflow.
Setting time makers against your project and its tasks is the definition of project time
management, regardless of the type of project. It entails assigning resources to each step and
specifying the time value of each activity.
Importance of Project time management,
One of the most crucial areas of the project management expertise is project time management,
which has a direct influence on a project’s quality, scope, and cost. Effective time management
ensures that projects are completed on schedule and within budget. It also makes clear.
• How long a project takes?
• which internal and external stakeholders should be included?
• When should their knowledge be included?
This procedure offers a structure for creating a list of tasks, their durations, resource estimates,
and how each fit into the project management plan as a whole.
When you know what you want to deliver and why, time planning makes it easier to get there.
This makes the time management phase of project management a crucial stage. It assigns a
value to every to every previously abstract subtask.

75 Thanoj E127966
Benefits of time management,
• Project time management might help you feel less stressed.

• Use project time management to boost output.

• Reduce project errors by using time planning

• Meeting lead time targets and avoiding last-minute rushes without compromising
quality are made possible by effective project time management.

• Time management for projects aids in goal achievement.

This is even more vital when it comes to implementing AI on the e-commerce platform of
Daraz and using the limited time and resources available. For the task to be completed within
the calculated time frame and the meetings to be scheduled as preferred, effective time
management practices will be necessary for the project to be a success.
The overall activities of data collection, system integration, model development, testing and
staff training will be organized by developing a detailed project schedule to maximize resource
utilization. Since it is the project manager’s task, the timeline will cover these key activities,
and address potential delays and adjustments to plans accordingly for the Daraz AI integration
project.
If there are factors such as system compatibility, staffing challenges, and minimizing
disruptions then effective time management at Daraz should be exercised. Having meeting
deadlines and continuous progress towards the introgation of the AI project gives a guarantee
of devoted stakeholder satisfaction and project credibility and possibility of success.

1.4.1 TimeLine for the Daraz E – commerce AI integration project

TimeLine,

Phase Task Name Start Date End Date Duration

Initiation Acceptance of the 2025.01.12 2025.01.18 5 Days

project proposal

Stakeholder’s 2025.01.17 2025.01.21 5 Days

Identification

Project Charter and 2025.01.22 2025.01.22 1 Day

Development of the
scope

76 Thanoj E127966
Planning Collecting Requirements 2025.01.23 2025.02.04 13 Days

Assessment of AI 2025.02.05 2025.02.15 11 Days

technologies

Making a project 2025.02.16 2025.02.16 1 Day

schedule
A strategy for Risk 2025.02.17 2025.02.22 6 Days
Management
Planning a Budget 2025.02.23 2025.02.25 3 Days

Execution Design of the system 2025.02.26 2025.03.11 14 Days

Development of AI 2025.03.12 2025.04.09 4 weeks

Models

Integration of Systems 2025.04.10 2025.04.20 11 Days

Instruction and Records 2025.04.21 2025.05.03 13 Days

Control Observation and 2025.05.04 2025.05.17 14 Days

Reporting
Assurance of Quality 2025.05.18 2025.05.31 14 Days

User Acceptance Testing 2025.06.01 2025.06.07 7 Days

(UAT)

Closure Evaluation of the Project 2025.06.08 2025.06.14 7 Days

Closure of the Final 2025.06.15 2025.06.18 4 Days

Project
Table 5.Timeline of the project

77 Thanoj E127966
1.4.2 Examining the Time Plan

1. Initiation phase

• Overview of the Task - This opens the countdown to the realization of the project and
is part of the initiation phase as we obtain the approval and how to recognize
stakeholders and create the Project Charter.

Key tasks of this phase,

▪ Project Proposal Approval (5 Days): For Alignment with goals and Resources,
it is needed to get approval from key stakeholders.

▪ Identifying Stakeholders (5 Days): Identify all the stakeholders such as vendors,

developers and key team members.

▪ Project Charter and Scope (1 Day): The charter formalises the objectives and
scope ensuring alignment of the project against organisational goals.

2. Planning phase

• Overview of the Task - Requirements gathering, AI technology evaluation, scheduling

and budget are included on the list of things included in this phase.

Key tasks of this phase,

▪ Gathering Requirements (13 Days): It encompasses collecting functional and

non-functional requirements for the AI integration into the Daraz e commerce
system.

▪ AI Technologies (11 days): Researching the AI tools and evaluating if they fit
into the infrastructure of the platform.

▪ Developing Project Schedule (1 Day): Creating a project timeline comprising

all the milestones and deadlines.

▪ Identify foreseeable risks such as data privacy and platform compatibility and
plan the mitigations (6 Days Risk Management Strategy).

▪ Financial Management (3 Days): Ensuring that the money appropriated to the

project will be enough to procure all the necessary software, hardware, and the
human resources.

78 Thanoj E127966
3. Execution phase

• Overview of the Task - System design, model development, integration and training are
all activities of this phase.

Key tasks in this phase,

▪ Design the system (14 Days): A design of architecture to integrate AI into

Daraz’s existing platform.

▪ Development of AI Model (4 Weeks): Suggestions as well as automated

logistics

▪ System integratiojn (11 Days): Prepared internal infrastructure for seamless

operation of the AI system with the e-commerce infrastructure of Daraz.

▪ Training and Documentation (13 Days): Provide training to the teams

mandatory or supporting, and document well so that after their participation the
system remains sustainable for a long time.

4. Control phase

• Overview of the task - We monitor, quality assure, and test users for the system to meet
requirements.

Key tasks of this phase,

▪ Monitoring and Reporting (14 Days): Periodic checkups to check if the project
deviates and proceeds in the right direction.

▪ Testing the AI system for functionality, evaluation and performance (14 Days):
Follows a Quality Assurance (14 Days).

▪ User Acceptance Testing (UAT) (7 Days): This is User feedback to confirm the
system to be used or for the intended use also.

5. Closure phase

• Overview of the task - The last phase of closure reviews the outcomes, documents
learnings and closes the project.

Key tasks in this phase,

▪ Success Assessment (7 Days): Encodings project success based on individual

‘pro’s’ and ‘con’s’, as well as whether the project was a success or failure.

79 Thanoj E127966
 ▪ Closing the Project (4 Days): Declaring ending to the project and handing the
AI system to Daraz’s operation team for active management.

1.4.3 Gantt chart

Projects often involve a number of tasks and activities. When managing a complicated project
as a project professional, it can be challenging to keep track of the tasks, their relationships,
duration, and progress. Understanding how the various activities relate to one another is crucial
since a job can have an impact on how a project develops.
We can better manage the project timeframes with the use of a Gantt chart. To help you
effectively oversee the execution of your projects.
Project, program, and portfolio managers may easily create project plans by using a Gantt
chart, which essentially puts project activities on a visual schedule. Among many other
applications, Gantt charts are used in resource management, project planning, project
scheduling, and project tracking.
• Determine dependencies, divide the project into smaller pieces, and distribute resources
wisely.

• Establish a project timetable by deciding on the order and length of the task.

• Track the development of tasks, evaluate scheduled and actual schedules, and spot any
problems or delays.

• Communicate project plans and development in an intelligible manner to clients, team

members, and stakeholders.

• Make sure that everyone on the team is informed of their roles and due to sates, since
this will promote cooperation and coordination throughout the project.
(Anon., 2024)

As part of its AI integration project in the e-commerce platform, Daraz uses Gantt chart
technique as a core part of them. It is this strategic tool that helps project managers monitor the
project progress by means of visual representation of tasks, timelines, and milestones thereby
ensuring the project progress is efficient. As Daraz’s platform is highly specific, the Gantt chart
breaks the project into dedicated tasks such as AI model development, system integration, and
staff training. It lets the team know what key deadlines there are and it helps to spot any possible
hitches or delays before long.
A Gantt chart also provides a breakdown of task dependencies that are essential to understand
the sequence of action to be performed. The chart helps map out how each task depends and
connects to other tasks in order for resources to be allocated appropriately so that the tasks

80 Thanoj E127966
don’t result in overburdening only on some while at the same time achieving optimal workflow
efficiency. This clarity makes sure that each one of these stages flow then is executed within
the allowed time. What’s more, it helps in making better decisions because project managers
can change their plan in accordance with the changes in circumstances and change the schedule
of the project without it affecting the overall project schedule.
Additionally, the Gantt chart also has an important part in breaking the sterile atmosphere
running within the project to eliminate communications with them all project stakeholders.
Offering a concise and graphic depiction of the project plan and goals serves to keep everyone
— from AI specialists to operational staff — aligned as to everyone’s responsibilities and the
project goals. Collaboration and accuracy is fostered by the transparency of this process, to see
that each part of the project is on track is important. Moreover, the Gantt chart is beneficial in
minimizing the risks associated with the delivery of the project since it helps in determining
potential bottlenecks and critical paths that are necessary for timely delivery of the project. An
overall approach to AI integration for Daraz means that the integration of AI is more effective
and efficient and that finally Daraz’s AI initiatives are aligned to the strategic objectives for the
long-term growth and sustainability of the Company.

81 Thanoj E127966
Gantt chart of the project,

Table 6.Timeline

Table 7.Gantt chart

82 Thanoj E127966
1.5 Risk Analysis

The process of locating and evaluating any problems that can have a detrimental effect on
important business efforts or projects is knows as risk analysis. The purpose of this procedure
is to assist companies in avoiding or reducing such risks.
When doing a risk analysis, one must take into account the potential for unfavourable
occurrences brought on by either intentional or unintentional human activity or by natural
phenomena such powerful storms, earthquakes, or floods. Determining the likelihood of these
occurrences occurring and the possible harm they may cause is a crucial component of risk
analysis.

Figure 14.Risk matrix

Importance of risk management,

An organization’s flexibility and long-term profitability are improved by risk analysis, which
offers an organized method for evaluation uncertainties.
Risk analysis used by businesses and other organizations for the following purposes:
• Anticipate negative outcomes from unfavourable occurrences and lessen their impact.

• Assess if a project’s possible hazards are outweighed by its advantages; this helps with
decision-making when deciding whether to proceed with the project.

• Prepare for equipment or technology failure or loss due to unfavourable natural or man-
made occurrences.

83 Thanoj E127966
 • Effectively distribute resources, including personnel, funds, and time, where they are
most required.
(Kinza Yasar, 2023)

1.5.1 Risk Analysis of Daraz’s AI implementation project

Introducing AI in Daraz’s e-commerce competitive operations has huge opportunities such as

enhancing customer interactions and logistics. On one hand it offers a few challenges but also
several that have to be handled carefully. A breakdown below of the main risks and how they
can be mitigated is as follows.
1. Ethical and Compliance Concerns
AI adoption at Daraz is no different — because it’s ruled by such complex legal and ethical
considerations that they are essential to discuss. Headlining is of data privacy and fairness at
the consumer’s lens, both which are legally required from the standards of an AI system to be
compliant and operationally ethical.
Possible Challenges,
• There may be legal repercussions for breaking restrictions.
The violations of regulations by AI driven processes could involve significant legal problems
in regard of data privacy laws or consumer protection.
• AI decisions might unintentionally foster prejudice, which would compromise fairness.
AI models may unknowingly replicate bias, that could harm product recommendations,
customer service, or even hire processes from their trust in the platform.
• Some automated pricing techniques can be considered immoral.
Specifically, some of the AI enabled dynamic pricing strategies could be perceived as unfair
and exploitative if they exploit customers’ purchasing behaviors to the extent that the latter are
seen as unethical.

Solutions,
• Regular audits to ensure AI practices adhere to international and national regulations.
Aim to have regular audits on check the adequacy of AI practices relative to international and
national regulations related to data privacy and consumer protection.
• Use fairness tests to eliminate biases in AI decision making.
Fairness test AI models regularly to make sure product recommendations or customer service
are not being skewed by any bias that can skew treatment of our users unequally.

84 Thanoj E127966
 • Maintain transparency in pricing and recommendation driven by AI.
Customers of an AI model must be kept in the loop about how often decisions such as pricing
and product recommendations are done with the use of AI and how the AI influences them,
which should be transparent and justifiable to prevent ethical concerns.

2. Security and Data Vulnerabilities

Large data sets are required by AI, and for that security is always extremely important at Daraz.
Customers’ information may be exposed to cyber threats due to a weak security system, which
highlights the need for strong data protection measures within the company.
Possible Challenges,
• Critical data may be leaked by cyberattacks that target AI-driven operations.
Large amounts of sensitive customer data are taken care of by AI enabled operations. Without
strong security systems in place, this important data is more vulnerable to be cyber attacked
and breached.
• Unauthorized access to AI algorithms may lead to data manipulation.
Daraz is relying on AI algorithms to run the important business operation of the company.
Unnecessary instances allow unauthorized parties to manipulate the AI models, which results
in the wrong predictions or fraud, making business decisions and customer trust.
• Over reliance on AI might lead to judgmental errors in the system.
Although AI can increase the efficiency, over reliance on AI systems also has the potential of
making errors in the judgment. There are situations that may not be accounted for by AI, or in
general represented correctly by the human intuition, which might affect the accuracy of the
decisions.

Solutions,
• Use advanced encryption and firewall protection to strengthen cybersecurity
infrastructure.
This way, Daraz can fortify its cybersecurity infrastructure with advanced encryption as well
as robust firewall protection to minimize the odds of getting its sensitive customer data
breached or warded off by cyberattacks.
• Regularly update AI models and security protocols.
To continue to thump remotely, Daraz should make sure its AI models are updated regularly
and keep in step with the latest security protocols to avoid perpetuating threats and preventing
its overall accuracy, system reliability, and create new threats of staying ahead.

85 Thanoj E127966
 • Use multi factor authentication to prevent unauthorized access.
Daraz can use multi factor authentication to prevent unauthorized access by users accessing AI
algorithms such that only authorized personnel are in charge of sensitive systems and data.

3. Uncertainities in Finance and Investment

The financial commitment for implementing AI at Daraz is significant. If not planned and
managed correctly, there is a risk of cost overruns without delivering substantial ROI. Proper
budgeting, resource allocation, and risk management are crucial to ensuring the investment in
AI leads to long-term business benefits such as improved operational efficiency, enhanced
customer experience, and increased profitability.
Possible challenges,
• Substantial initial expenditures for AI technologies and skilled workers.
Upfront costs for starting to use AI technologies, software licensing, and paying people to run
and apply the AI solutions.
• Maintenance and updates may always expensive.
Recurring expenses may come from ongoing maintenance and system updates, and therefore
the Daraz’s AI infrastructure’s long term budget.
• AI may take some time to produce measurable financial benefits.
However, AI needs time to generate significant financial benefits as these systems would
require to be optimized and accepted into Daraz's working conditions for a better performance.

Solutions,
• Make through financial preparations before utilizing AI.
Daraz should thoroughly make arrangements for the financial side of AI before implementing
it to avoid overspending or paying for something not expected.
• AI will be used in phases to balance benefits and costs.
Daraz will introduce AI gradually at Daraz to strike a perfect balance between benefits obtained
and costs incurred as a result to make sure that it is felt smoothly and adopted appropriately.
• Monitor AI’s performance to ensure that it contributes to long-term profitability.
Monitor AI’s performance constantly to ensure that it enhances both Daraz’s operational
efficiency and customer satisfaction as well as long term profitability.

86 Thanoj E127966
 4. Disruptions to Operations and integration
The main focus is to integrate AI with Daraz’s existing systems like order management,
inventory tracking and customer service without disrupting existing operation. Lack of
integration could result in major operational disruption resulting in dissatisfaction to customers
and inefficiency of business processes.
Possible challenges,
• Compatibility problems with older systems and AI.
The integration of new AI technologies with older systems is riskily incompatible, and may
cause some challenges.
• Revenues and customer experience may be impacted by outages during AI adoption.
In the case of AI adoption, there is a possibility of system outages due to which the revenue
generating capabilities kitchen and customer experience may be adversely affected.
• New AI powered procedures may be met with resistance from employee
The adoption of new AI powered procedures raises the issue of employees resistance to
embrace such procedures.

Solution,
• Before implementing AI on a large scale, conduct small scale pilot testing.
Daraz tests the efficacy of AI small scale and finds any gaps in the process before it’s fully
rolled out.
• Offer training programs to help employee get used to AI powered processes
To accommodate the employees with AI powered processes and make them use them
efficiently, Daraz has given training programs which are all general.
• Maintenance your hybrid strategy, where AI supports human decision making rather
than replaces it.
At Daraz, we have adopted a hybrid strategy in which AI strengthens the decision of a human
with the aim to make the automation much better than the human involvement.

5. AI Mistakes and problems with Customer Trust

Depending on how accurate your AI is, the efficiency of your AI comes down to it, however,
errors in AI driven recommendations or pricing models can result in negative customer
satisfaction.

87 Thanoj E127966
Possible challenges,
• AI genarated errors might lead to incorrect product suggestions.
AI mistakes by Daraz can make wrong product suggestions being offered trending products to
the people, which can impact customer satisfaction and sales as result.
• Excessive customisation may make customers uneasy.
Customers might find excessive customization driven by AI gelling them out of touch, as they
might sense that their preferences are being over manipulated.
• Without human oversight algorithmic biases might be unintentional.
The presence of algorithmic bias/algorithms in Daraz AI models is possible without human
supervision, thus resulting in cases of unintentional discrimination or recommending
something that is unfair.

Solutions,
• Create a feedback loop so that users may report AI errors.
So, in Daraz create a feedback loop that users can easily report AI error to improve system
accuracy and efficiency.
• Provide succinct explanations for AI driven decisions to maintain transparency.
Giving concise explanations for AI generated decisions to remain transparent and promote trust
among customers, they need to have the rationale of the recommendation.
• Ensure that people may access AI driven processes.
At Daraz, make sure that people have easy access of AI driven processes so that employees and
customers can use these AI tools easier and faster.

With all the potential that AI will bring in the future operations of Daraz, the company can
quickly face major challenges if not taken care with caution. With a proactive approach to
eliminating risks with robust strategy, Daraz can successfully implement AI. This means
protecting the customer, protecting data security and AI optimization for the long term
profitability. With this, Daraz can fully reap the benefits of the power of aligning AI, while
upholding trust, security and continuing to stay competitive in the market.

88 Thanoj E127966
1.6 Quality Plan

A quality management plan is a document that aids in the execution of quality assuarance and
management tasks by the project manager and team. In project management, quality refers to
meeting the demands of the client and the project’s specifications.
A quality management plan is part of an oerall projaeact plan and outlines the actoins you will
take to satisfy the project’s quality goals over the course of its life cycle. These actions (as well
as the tools required to carry them out) are also detailed in the quality management plan.
Expecying quality without making ant preparations is a stupid idea. Quality is deliberate and
demands deft execution. The first step in establishing and formalizing the procedures required
to meet the project’s quality standards is creating a quality management plan. Project
management software that can arrange and distribute the plan to the project team is the most
effective tool for this.
Why Is a Quality Management Plan on a Project Important?
A quality management plan is a crucial part of delivering successful project for your
stakeholders, but it works for more than just one project. Customers expect consistency, and
you can consistently meet their quality expectations by having a plan in place to maintain a
quality production for whatever product or service you’re manufacturing. Quality management
is also not just for the project team, when everyone in the organization understands the plan,
efficiency increases, while it’s crucial that everyone involved in the manufacturing process uses
the quality management plan as a guide, having all departments familiar with the plan increases
efficiency and establishes a common goal that everyone can understand.
Customers will notice when you produce high quality goods or services. Your brand stands out
in the competitive market thank to its quality. Delivering consistent quality is made easier with
the aid of a quality management strategy. Customers get devoted to your goods or service as a
result. Profitability results from brand value, which is derived from loyal consumers.
(Malsam, 2022)

To deploy AI successfully to Daraz, AI needs to be seamless integrated within its operations

within an overall Quality Plan (QP). This plan is to make sure that AI driven solution follows
the goals of the business as well as comply with industry regulations and standards. The QP
enhances system reliability, optimizes data accuracy, and acting in an ethical manner,
guarantees the use of AI. It finally results in operational efficiency, better customer experience
and cheaper e-commerce solutions that help Daraz to remain at the top of the game and earn
value for the customers and stakeholders alike.
The tenets of the QP for this prioritization are accuracy of the AI model, ethical compliance,
and consistent performance. To test the product recommendations, fraud detection and
automated customer support, rigorous testing protocols are implemented and the algorithm is
validated comprehensively. Regulations meant for protecting consumers' lifeline will be
followed by transparent AI governance that also reduces bias and ensures fairness. This

89 Thanoj E127966
approach ensures the reliability, ethics, and adherence to the business goals as well as the
consumer rights of Daraz’s AI solutions.
Using consistent and accurate AI generated insights, the users will get enhanced satisfaction
and trust. For personalized shopping, inventory management and consumer trend experiences,
high quality data should be a parity. Open communication about AI’s capabilities and
limitations fosters confidence among customers and stakeholders. The AI system is
continuously monitored and if its execution is not optimal, the AI system adapts to the changing
data patterns and keeps maintaining its relevance in Daraz’s operations. By following this
approach, the customer loyalty would be reinforced, decisions would be made better, and
operational efficiency would be increased.
The Components of quality management
• Reliability, accuracy, and adherence to ethics

• Improved Trust and Customer Experience

• Operational Efficiency and Cost Control

• Cooperation and Education of Employees.

In addition it emphasizes that Daraz must during the deployment of AI minimize operational
inefficiencies and control costs. Quality control mechanisms are built in to help maintain the
data accuracy and perform the process of any beginning machine learning more productively.
The logistics using the Predictive Analytics happens to save you the unnecessary costs and also
enable automating order fulfillment as well as supply chain operations. It results in a
streamlined and economical process where Daraz can meet the customer demand while being
profitable and operationally excellent.
It is necessary for AI to succeed, through employee engagement. IT, customer service and
supply chain teams maintain a cross departmental collaboration to make sure that the AI
solutions they deploy serve the business objective. It pre processes and validates the protocol,
whereas preventive measures for data integrity assure there are no inconsistencies and carry
stronger safeguard to data driven decision making. Thanks to the well structured quality plan,
Daraz has also gained high standards and its reputation among new customers and partners as
well has supported business growth through its cutting edge AI solutions. The use of this
approach enables Daraz to stay at the top of the competitive e-commerce world.

1.6.1 The AI quality Control Framework of Daraz

Daraz will need a robust quality control mechanism to implement AI, for its accuracy and
efficiency, and to maintain ethical compliance. To run at high performance in these operations
driven by AI, these mechanisms are critical to gaining the customer trust and regulatory
adherence required. the quality control measures include:

90 Thanoj E127966
 1. Management of Data integrity,
High quality, validated, and up to date data is the key to the correct working of these AI models
at Daraz. An important practice to develop robust data governance is to make sure errors don’t
befall in critical areas like product recommendations, fraud detection, and logistics
optimization. For accuracy and relevance, Daraz should go ahead and regularly audit its
datasets, leverage various data sources, and real time updates should be ensured. Almost every
aspect of these practices will significantly improve the performance of AI, thus making it more
trustworthy and reactive to customer needs while reducing the possibility of wrong decisions
or stoppages of planned activities.
2. Algorithm Openness
Daraz’s stakeholder trust is fostered thanks to AI systems being transparent. This is crucial and
can be done by giving insight into AI decision making making processes, like how algorithms
look at products, how they spot fraud or even personalise user experience. Daraz employs
explainable AI models and provides a holistic formulation of documentation to propagate
ethically plausible AI utilization, decrease biases and make sure the AI led decisions align with
both the customer’s expectation and the industry reference. The transparency part is key to
building a relation with customers and partners based on credibility and long term.
3. Ongoing system Improvement
Daraz’s AI models continue to work due to constant updates and continuous improvement.
Such systems need to adapt to the ever changing consumer’s preferences, industry trends and
business requirements. By operating on feedback received from system evaluation in any given
environment, the AI systems continue to fine tune themselves to serve a variety of purposes
and fulfill their mission. Daraz can incorporate real time data and insights to bring the value of
its AI driven solutions to customers and improve operational performance.
4. Strategies for Risk Mitigation
Algorithmic biases, data security breaches and incorrect predictions are the major risks that
may occur on the AI-driven e-commerce at Daraz. Daraz must implement some robust
cybersecurity practices, perform fairness tests on its AI algorithms and have contingency plans
in place to avoid disruptions to the business and smooth the AI operations. Addressing these
potential risks proactively will allow Daraz to keep customer trust, regulate within regulatory
standards and maintain the reliability of accuracy of its AI-powered services.
5. Metrics for Performance Evaluation.
To assess the efficiency of AI at Daraz, the development of key performance indicators (KPIs)
is essential. To measure AI’s impact, the company uses measurement KPIs such as conversion
rate, fraud detection accuracy, recommendation efficacy and delivery optimization. Daraz can
continuously evaluate these metrics to increase its AI driven and optimize for better business
outcomes as well as better customer experience.
6. Respect for the law and Compliance
Daraz, being internet based purchasing, needs to ensure ethical AI, GDPR data privacy laws as
well as consumer protection regulations applied to its AI in it. Without such audits and
continuous monitoring of changing regulatory standards, the AI systems will be no longer
91 Thanoj E127966
compliant, reliable and trustworthy. Such proactive approach ensures that Daraz protects
customer’s data, remains legally compliant and at the same time upholds its reputation as a
trusted e commerce platform.
7. Implementing AI ethically
Daraz should use the AI responsibly so that it doesn’t lead to the unethical data exploitation,
unfair pricing, and biased product exposure. Daraz sets up clear ethical criteria in which the
business practices AI driven decisions to be a conform relationship with those coined by the
brand upholding their reputation and the trust of its customers. It aids Daraz in its striving to
uphold the standard of integrity while improving the experience of customers and operational
effectiveness.

1.6.2 Quality Plan for AI in E – Commerce at Daraz Sri Lanka

To stay out of the ethical basement, Daraz has to manage its AI really smartly so that the
company does not exploit unethical data, sell at unfair prices and suggest biased products to
consumers. Daraz sets clear ethical boundaries that the decisions made using AI should abide
by, strictly following industry’s best practices and the laws of the land. This is good in
protecting Daraz’s brand reputation as well as guaranteeing that customers get a fair deal and
won’t have to put up with a warlord installation on their wallet, and that it’ll work with both
customers and partners. At the same time, it helps keep the AI powered solutions in line with
Daraz’s long term business objectives and regulatory constraints, leaving no doubts that it stays
true to its innovation and customer satisfaction promise.
1. Adherence to Regulations
To maintain full compliance with global and local regulations, AI driven customer interactions
and data storing and payments processing are done on their platform to enhance security and
operational integrity.
Suitable Standards,
• GDPR – Safeguard privacy and security while customizing using AI.

• ISO 27018 – Protects customer information kept on cloud servers.

• Sri lanka ICTA Guidelines – Aligns the application of AI with national e – commerce
regulations.

2. Tracking the performance of AI

Always, Daraz is continuously monitoring AI performance to improve system efficiency,
operationally optimize and increase the platform effectiveness..
Suitable Standards,

92 Thanoj E127966
 • ISO 9001 – Evaluate AI’s effectiveness in search rankings, logistics, and fraud
detection.

• Balanced Scorecard – Assess the impact of AI on e – commerce sales and consumer

engagement

• KPI’s for AI – Monitors AI driven conversion rates, fraud prevention, supply chain
accuracy.

• Lean Metrics – Assures that AI lowers waste and boosts logistical efficiency.

3. Moral AI Procedures
The pricing, recommendation and decision making with Daraz are on a more ethical path and
based on transparency, accountability and fairness.
Suitable standards,
• OECD AI Principles – Fosters ethical AI development and builds customer confidence.

• IEEE Ethically Aligned Desing – Ensures AI emplies with moral standards and ethical
online business practices.

• ACM Code of Ethics – Ensures that AI upholds fair competition and uset privacy.

4. Risk Management for AI

AI risk management procedure in Daraz is robust, which renders errors, biases and, security
vulnerabilities, for a safe and dependable system.
Suitable standards,
• ISO 31000 – Draws attention to the dangers AI presents to inventories, price, and
customer pleasure.

• NIST SP 800-30 – Strengthens AI to identify fraud in the face of cyber threats.

• COSO ERM Framework – Offers strategies for lowering AI risks across the board or
the whole organization.

• ISO 14971 - Lessens the risks associated with AI’s use in vendor management and
logistics

93 Thanoj E127966
 5. AI data Governance and Quality
At Daraz, accurate, secure and well maintained data is the main foundation of building effective
AI applications.
Suitable standards,
• ISO 8000 – Improve data accuracy for customization driven by AI.

• ISO 9001 – Arranges for all system’s data handling

• DAMA-DMBOK – Ensures that AI training data is sourced and managed ethically.

• GDPR Compliance – Protects customer data while utilizing AI to make judgments.

6. Algorithm Openness
Daraz ensures algorithm transparency to ensure that the sellers, consumers, and the regulatory
authorities get the benefit of a fair and responsible AI decision making.
Suitable standards,
• ISO/IEC 27001 – Defends AI powered ranking and pricing systems against online
attacks.

• IEEE P7001 – Guarantees the transparency of AI system suggestions and search results.

• ISO 25010 – Evaluate the reliability and quality of AI systems.

• NIST AI Risk Management Framework – Expands AI’s role in Pricing and fraud
detection techniques.

7. AI’s Continuous Improvement

As they are always evolving customer preferences, technological advancements and market
dynamics, we need to adjust our AI models in Daraz to stay competitive.
Suitable standards,
• ISO 9001 – Encourage continuous improvements in AI models.

• Six Sigma – Reduces inefficiencies in AI driven logistics and consumer analytics.

• PDCA Cycle – TO enhance AI, employ a systematic Plan-Do-Ckeck-Act methodology.

• TQM (Total Quality Management) – Encourages the development of AI to improve e

commerce in general.

94 Thanoj E127966
Daraz integrates AI with a robust quality framework to ensure its operations in Sri Lanka align
with global standards, ethical guidelines, and business objectives—enhancing customer
satisfaction and operational efficiency.

1.6.3 Quality Control Strategy for Daraz Sri lanka’s AI integration

To stay out of the ethical basement, Daraz has to manage its AI really smartly so that the
company does not exploit unethical data, sell at unfair prices and suggest biased products to
consumers. Daraz sets clear ethical boundaries that the decisions made using AI should abide
by, strictly following industry’s best practices and the laws of the land. This is good in
protecting Daraz’s brand reputation as well as guaranteeing that customers get a fair deal and
won’t have to put up with a warlord installation on their wallet, and that it’ll work with both
customers and partners. At the same time, it helps keep the AI powered solutions in line with
Daraz’s long term business objectives and regulatory constraints, leaving no doubts that it stays
true to its innovation and customer satisfaction promise.
QMP (Quality Management Plan) gives clear understanding of how AI can improve business
efficiency but at the same time reduce the risk of biased algorithms, poor recommendations,
and security issues. Below are the key aspects that illustrate the need for Quality management
in AI integration for Daraz Sri Lanka.
1. Improving AI accuracy and Decision Making
The importance of Accuracy in AI systems
After the implementation Daraz uses AI-driven algorithms in product recommendations, fraud,
auto customer services, and logistics optimization to make things efficient and satisfying. The
AI-driven features are made to provide accurate outputs that ensure smooth workings and
improved conversions. AI algorithms have the capability to produce bugs if not handled
properly, and this can lead to poor recommendations, inefficient fraud detection optimization,
decision-making, and low customer satisfaction. These vulnerabilities have a direct effect on a
company's performance, and thus Daraz must continuously improve AI models, implement
quality controls, and stay open to continue enjoying confidence and operational dominance.
Keeping Accuracy Through Quality Control.
To maintain Precision, Daraz Sri Lanka must,
• Using robust data validation techniques will ensure that Daraz's AI models are trained
on real and high-quality data sets, minimizing errors while suggesting products,
detecting fraud, and planning logistics.

• Real-time monitoring of AI algorithms provides Daraz an opportunity to notice

mistakes in real time so customers are suggested pertinent products, their suspicious
activities get detected successfully, and automated activity operates optimally.

95 Thanoj E127966
 • Regular performance testing and refreshing enable Daraz to improve AI-based
decision-making, user experience, and responsiveness to market trends while
maintaining accuracy and efficiency in its e-commerce activities over time.

Anticipated results,
• Enhanced customer retention and engagement with AI-powered personalized product
suggestions, allowing customers to find what they need in a timely manner and enhance
overall shopping experience.

• Decreased order processing errors with automated core functions like checkout,
payment validation, and inventory management, providing an even smoother buying
experience.

• Enhanced inventory management and logistics effectiveness as AI streamlines stock

management, demand planning, and delivery scheduling, minimizing delays and
reducing operating costs.

2. Building Trust with Customers and Transparency

Why Transparency of AI important?
To be able to use AI in its operations effectively, Daraz needs the trust of its customers. AI-
driven systems such as search algorithms, pricing, and personalized deals need to be transparent
and fair in order to retain credibility. Suspicion of bias or manipulation of prices or
recommendations by users would result in a loss of credibility, more complaints, and doubt
regarding the use of the platform. Transparency in AI-driven decisions and ethical AI practices
will enable Daraz to gain customers' long-term loyalty and trust.
Ways to increase Transparency in AI,
• Deploying Explainable AI (XAI) Frameworks: Customers need to have some idea of
how AI-driven decisions, i.e., product recommendations and dynamic pricing, are
arrived at. This builds trust and dispels suspicion.

• Instilling Fair and Unbiased Algorithm Training: AI algorithms need to be trained on a

variety of datasets to avoid bias in search rankings, price modification, and personalized
promotions so that there are level-playing fields for all sellers and equitable experiences
for buyers.

• Providing User Control Over AI-Generated Recommendations: Customers need to be

provided with choices to personalize or modify AI-generated recommendations,
allowing them to filter their shopping interests according to their requirements.

96 Thanoj E127966
Anticipated results,
• Increased Brand Image and Customer Trust: Through adopting transparent, unbiased,
and precise AI-based dynamic pricing and personalization systems, Daraz can increase
customer loyalty, consolidate its brand image, and reap consumer trust, resulting in an
increased rate of satisfied customers and repeat buying.

• A More Equal and Inclusive Shopping Experience: With impartial AI algorithms and
customized customer experiences, Daraz is able to provide all customers regardless of
demographics with equal access to relevant products and offers, leading to an inclusive
shopping experience.

• Less Complaints Regarding AI-Powered Suggestions or Prices: By being customer-

focused and more transparent in AI-powered processes, Daraz can reduce the number
of customer complaints regarding product suggestions or dynamic prices. Customers
will feel better about the fairness of AI mechanisms and complain less.

3. Ensuring Compliance with Rules

The role of compliance in AI governance,
For Daraz, it is essential that its AI technologies adhere to international and regional standards,
including Sri Lanka's ICTA guidelines, GDPR (for related parties), and ISO data protection
standards. Adherence to these regulatory guidelines guarantees that AI-driven operations,
including personal recommendations and fraud prevention, are secure and trustworthy. Non-
compliance might result in very harsh legal repercussions, loss of information, or reputation
damage, which may, in turn, impact customer confidence and business. Good governance is
thus critical in ensuring the sanctity of AI systems and keeping Daraz's reputation intact.
Maintaining Daraz Sri lanka AI compliant,
• Enforcement of Ethical AI Guidelines: Daraz should implement ethical AI practices
that ensure compliance at a worldwide level to bring about accountability, transparency,
and equity in the decision-making done through AI.

• Improving Data Protection: Data protection becomes inevitable by offering a tight

coverup for safeguarding personal data, ensuring customers' faith in the portal.

• Constructing AI Audit Trails: A record of the activity and thought of the AI system must
be made in order to provide an audit with an open history, assure compliance with law,
and enable detection of any possible problem with the system simpler.
Anticipated results,
• Avoidance of Legal Charges and Criticism: Compliance with the regulatory norms
and ethical use of AI practices by Daraz Sri Lanka will prevent them from potential
legal charges and penalties, providing hassle-free operations on local and global
platforms.

97 Thanoj E127966
 • Heightened Investor and Customer Trust: Heightened investor and customer trust due
to regulation compliance of AI along with information utilization at an ethical level,
ultimately resulting in improved brand image and business progression.

• Improved Data Security and Ethical AI Use: By focusing on data security and ethical
AI use, Daraz can provide a safe environment for customers with reduced risks and a
positive image of the company's dedication to privacy and justice.

4. Cutting down on AI risks and errors.

Possible Dangers of AI in Online purchases
If not controlled, AI processes utilized for pricing, product visibility, fraud detection, and order
fulfillment lead to severe problems, including mispricing, biased product visibility, false fraud
signals, and order fulfillment problems. These errors have significant implications on sales
performance, customer satisfaction, and brand trust, hence compromising the entire customer
experience on Daraz. Hence, constant monitoring and refining of AI processes are required to
reduce risks and build a satisfactory shopping experience.
Techniques for Quality control to Lower AI Risks,
• Automated Error and Testing Detection: Automated systems continuously monitoring
AI-based processes assists in detecting errors or inconsistencies at the right time. This
maintains the accuracy of AI systems and the optimum level of efficiency and reduces
downtime at Daraz.

• AI Decision-Making under Human Supervision: Integrating AI-based decisions with

human controls ensures key business activities such as pricing, suggestions on products
to purchase and sell and order fulfillment remain in sync with Daraz business strategy
and consumer aspirations. Human controls curtail risks of AI-based errors and
prejudice.

• Routine Update of AI Models:It helps in regularly refreshing AI models in order to

adjust to emerging technology developments and shifting consumer patterns. This
enables Daraz to maintain up-to-date and relevant AI predictions and thus an
uninterrupted and up-to-date customer experience.

Anticipated results,
• Less Customer Complaints: With AI-informed correct product recommendations and
pricing, inappropriateness-related and misadvice-related complaints by customers are
lessened, leading to better customer satisfaction.

• Improved Fraud Detection Precision: AI-based solutions will result in fewer false
positive alarms and improve fraud detection precision and accuracy, minimizing the
risk of flagging legitimate transactions and building confidence in the platform.

98 Thanoj E127966
 • Enhanced Inventory and Logistics Control: AI-based systems will better track inventory
and manage logistics activities to improve order fulfillment efficiency and reduce
stockouts and other supply chain inefficiencies.

5. Enhancing AI system’s scalability and performance

The role of AI in scalability in Business,
In order to enable Daraz to expand in Sri Lanka, it is required that its AI order fulfillment
technologies, chatbots, and data analytics are optimized to handle increasing demand. Without
this, these AI technologies would result in slow site response times, order fulfillment mistakes,
and site crashes during peak traffic, to the loss of customer satisfaction and business
productivity. Frictionless AI integration is required to enable efficiency on an ongoing basis,
improve consumer satisfaction, and enable long-term business expansion.
Quality Management Techniques for AI scalability,
• AI Performance Under Heavy Usage: Daraz conducts constant load tests to study AI
behavior in times of heavy traffic to ensure there is proper functionality in times of
heavy use.

• AI Cloud-Based Solutions: Daraz employs cloud-based AI solutions, which provide the

flexibility to scale the resources quickly and ensure consistent performance during the
highs and lows in demand.

• Ongoing AI model creation and improvements: Daraz regularly creates and improves
its AI models to refine them so that the system gets better at keeping up with changing
consumer requirements and new market trends and attains incremental business
outcomes.

Anticipated results,
• Effortless Sales-Rich Periods Management: With AI-based solutions, Daraz can easily
manage sales-rich periods including festive sales and Black Friday and maintain a
smooth selling process and happy customers.

• Faster Response Time of AI-Powered Search Engines and Chatbots: AI optimization

will enable search engines and chatbots to reply with greater speed and accuracy,
leading to higher levels of customer satisfaction and involvement.

• Enhanced Inventory Forecasting and Supply Chain Optimization: AI helps in enhancing

the precision of inventory forecasting such that Daraz can better forecast demand,
optimize stock levels, and optimize the supply chain for increased efficiency.

99 Thanoj E127966
 6. Improving Security and Data Governance
The significant data security and quality
Daraz's AI relies on vast quantities of consumer, transactional, and product data to generate
accurate predictions for product, price, and logistics efficiency recommendations. If the AI
models themselves make incorrect predictions due to low quality data or security breaches,
consumer confidence is lost. This could cause incorrect product recommendations, incorrect
prices, or security breaches at the expense of the consumer experience and ultimately the Daraz
brand. Daraz needs high security and high-quality data in order to maintain customer
confidence and AI success.

The best methods for data governance,

• Strict data validation rules: Daraz should maintain stringent data validation rules in
order to use only accurate and reliable data for training its AI systems. This ensures the
product recommendations, pricing rules, and anti-fraud rules are correct and working
as anticipated in order to maintain satisfied customers.

• Encryption of confidential client information: Daraz protects its clients' information by

encrypting confidential data such as payment and personal information at the highest
level of security standards. This ensures data security and offers confidence to clients
who entrust their web-based transactions with Daraz.

• Routine AI security scanning: Daraz regularly scans the AI systems' security for
exploitable weaknesses by which they can be attacked. Through constant scanning of
the security structure of the AI, Daraz can avoid intrusions and maintain its systems as
secure as possible in order to protect user information and maintain business continuity.
Anticipated results,
• AI powered by accurate and reliable data: Daraz ensures its AI programs are powered
by accurate and quality data so that it offers a better user experience. Whether in the
shape of suggestions for products based on individual preferences, accurate pricing, or
optimized logistics, all AI-based decisions are made on reliable data, making the AI
more useful and reliable for business.

• Increased consumer confidence: With the more stringent data protection law in place,
Daraz can assure its buyers the safety of their transactional and personal information
being processed. This transparency and commitment to data protection instill
confidence in the consumer and play a key role in building and sustaining loyal buyers
and winning over new ones.

• Adherence to international norms of data security: Daraz follows international norms

of data security such as GDPR and ISO norms in order to remain in compliance and
avoid security violations. In this process, Daraz protects the information of the

100 Thanoj E127966

 customers, reduces risk, and achieves a good image as a safe e-commerce platform in
the global market.

7. Promoting Ethical AI Techniques

The Ethical challenges of AI,
The Ethical Challenges of AI: Daraz recognizes the importance of fairness in AI to prevent
ethical issues such as unfair product visibility, discriminatory pricing, or misleading
advertisements. By implementing ethical AI frameworks, Daraz ensures that its AI systems are
designed to treat all customers fairly, avoiding manipulation and bias. This commitment to
fairness helps maintain trust with consumers, aligns with ethical standards, and supports
Daraz's reputation as a responsible e-commerce platform.
Implementing Ethical AI at Daraz,
• Diverse Training Datasets for AI to Steer Clear of Recommendation Biases: Daraz
ensures their AI gets trained on representative and varied datasets to avoid biases in
recommendations. With an integrated array of data, Daraz provides balanced and
individualized shopping experiences to users.

• Simple and Clear AI Transparency for Consumers: Daraz prioritizes transparency in its
AI-based activities by keeping the consumer properly informed in plain language
regarding how the pricing and recommendations mechanism work. This builds
confidence and allows the consumer to feel comfortable with its AI-based interactions.

• Ethical Frameworks for Continuous Development: Daraz conducts periodic ethical AI

audits to review and improve its AI systems from time to time. These audits ensure AI
practices have as few biases as possible and as much fairness as possible and lead to
continuous improvement in business and customer satisfaction.

Anticipated results,
• Equitable Buying Choices for All Consumers: With ethical AI, Daraz offers all
consumers from any origin equal and impartial recommendations and prices so every
consumer gets an equal and inclusive buying opportunity.

• Enhanced Ethical Reputation for AI Use: Daraz's commitment to ethical AI maintains

its reputation as an organization dedicated to fairness and transparency. This promotes
consumer confidence and positions Daraz as an e-commerce industry leader with an
ethical approach to business.

• Higher Involvement from Ethically-Minded Consumers: Ethical AI attracts those

consumers who are concerned with sustainability, justice, and openness in buying and
selling practices. This enables Daraz to connect with a consumer segment interested in
ethical business at a deeper level.

101 Thanoj E127966

 8. Acquiring a Competitive Edge
How AI enhances market leadership,
Organizations possessing quality AI solutions such as Daraz are positioned head and shoulders
above their competitors in a number of key domains such as consumer engagement, logistics,
and sales maximization. With AI in improving personalized shopping, streamlining supply
chains, and building efficiency in business activities, Daraz builds up a sustainable competitive
advantage. Efficient AI management keeps Daraz at the pinnacle of market trends, responds to
consumer concerns in good time, and remains leading in the e-commerce sector. This strategic
deployment of AI allows Daraz to not just meet existing consumer needs but also to anticipate
and foresee upcoming needs and remain at the top in the market.
Crucial AI strategies to boost competitiveness,
• AI powered customization in marketing t increase conversion rates.
AI can be used to provide personalized product recommendations, targeted promotions, and
personalized marketing messages, thereby improving conversion rates and increasing sales for
Daraz. Personalized experiences create customer loyalty, leading to repeat purchases and long-
term growth.
• Utilize astute inventory forecasting techniques to reduce shortages of supplies.
By employing AI based analytics, Daraz can ascertain customer demand with the highesr level
of precision, lowering the risk of an inventory shortage and ensuring that popular items are
kept in stock. This increases customer satisfication and allows supply chain optimization
through timely demand fulfilment.
• Efficient fraud detection to lower financial risks.
Using AI for real-time fraud detection is a powerful tool in Daraz's arsenal against financial
threats. AI algorithms discern transaction patterns, identify unusual behavior, and actively
mitigate fraudulent activities. Thus, customers and the company will continue enjoying
protection from financial loss, bolstering trust in the platform with secure transactions.

Anticipated results,
• Higher client satisfaction and retention rates.
Better Customer Satisfaction and Retention: Daraz delivers personal experiences and
integrated AI engagements to please customers, keeping them retained for the long term and
retaining more of the nudge.
• Reduced operational costs as a result of AO powered automation.
Daraz employs the considerable potency of AI automation — in inventory management, order
fulfillment and customer support processes — resulting in an incredibly high savings on
operational costs.

102 Thanoj E127966

 • Enhanced brand distinctiveness and customer loyalty.
AI makes for a stronger Brand Identity: The creation of a separate brand presence and the trust
of customers that builds to a higher level of customer loyalty is made possible because AI gives
out consistent, reliable and specific service.

Daraz Sri Lanka boasts a strong Quality Management Plan for AI to meet all the terms of
reliability, security, ethics and efficiency of its AI solutions. However, Daraz’s continued
competitiveness is based on continuous improvement of accuracy, compliance, transparency,
risk management to achieve superior customer experience and operational excellence.

1.6.4 Quality Management Tools that selected for Daraz’s AI

implementation project

Fish Bone Diagram,

A fishbone diagram is a visual aid for classifying the possible causes of an issue. It is sometimes
referred to as an Ishikawa diagram or a cause and effect diagram. This rechnique is used to
determine the underlaying reasons of an issue. At its core, the fishbone diagram is a tool for
casual analysis that shows the causes and sub causes of a given problem or effect.
A fishbone diagram is a form of mind map template that blends brainstorming with the process
of root cause analysis. As a test case approach, it ought to be effective in identifying cause and
effect.
Usually used to center a discussion on an issue, a fishbone diagram is helpful in project
management , product development, and troubleshooting procedures. Numerous business make
use of fishbone diagrams. Fishbone diagrams, for example, are used in the healthcare,
education, and service sectors for a variety of objectives, including process optimization and
customer service problems.
Following the group’s brainstorming of every potential source of an issue, the facilitator assists
the group in ranking the causes in order of significance and creating a hierarchy. The diagram’s
design, which resembles the skeleton of a fish, is where the term originates. Usually, fishbone
designs are created from right to left, with each big “bone” of the fish branching out to form
smaller, more detailed bones.

103 Thanoj E127966

 Figure 15. Fishbone diagram

Daraz’s AI integration project according to Fishbone diagram

Main Problem – Daraz Sri lanka’s AI integration Challenges,

1. People, (Big bone)

• Worker’s inability to use AI (Sub bones)

As can be the case, some employees at Daraz may have difficulty adapting to AI process
because of knowledge or lack of experience in these technologies, resulting with workplace
background not yet adapted to when AI is smoothly integrated into routine.
• Resistance to the deployment of AI (Sub bones)
One of the possible reasons for this is that some teams at Daraz, especially those that have been
around for years, may have reservations over full scale deployment of AI since they fear for
their jobs, or concerned about workflows, being disrupted.

• Not enough training programs (Sub bones)

This may result in the gap between AI and employees at Daraz which will make it hard for
Daraz employees to upgrade their skills and use them through AI tools efficiently and reduce
the company’s ability in making maximum use of AI to improve its efficiency and services.

104 Thanoj E127966

 2. Process,

• Poor data management

This could mean problems with data quality and organization at Daraz that could put into
jeopardy the use cases where data can be utilized to generate insights by using AI models,
ultimately affecting more erroneous outcomes.
• Systems for quality control that are inadequate
Inability of robust monitoring and ensuring quality of AI processes at Daraz can affect the
performance and reliability of AI driven solutions at Daraz.
• Insufficient principals for AI governance
However, Daraz might face issues regarding the integration of it in broad AI governance
frameworks which would entail enforcing ethical use, transparency and compliance on
offerings involving AI.

3. Technology,

• The restricted ability of AI infrastructure to grow.

The scalability of its AI infrastructure becomes problematic for a Daraz, which will then
struggle to scale its business, adapt to more data and processing needs.
• Problems integrating with current systems.
It may prove to be difficult for Daraz to integrate AI in the already existing systems as it
could result in disruptions in operations or inefficiency during the transition of Daraz systems
to AI-based solutions.
• Expensive upkeep
In case of Daraz, the costs of maintaining and updating AI infra could become expensive as
technology progresses and it necessitated a regular change of such hardware so that it work
optimally and does not get baseline.

4. Data,

• Incomplete or inaccurate datasets.

An issue with completeness and accuracy of data can make it difficult for Daraz to formulate
effective and reliable AI models used for customer insights, pricing and inventory
management.
• Security and privacy issues.
Data protection is so important to Daraz that any lapse of data security or privacy would
smear compliance and customer trust as well since AI depends so heavily on very sensitive
information.

105 Thanoj E127966

 • AI decisions are influenced by biased training data.
Bias in training data can gain variations in the way the model trains – and the results of the
model can become unfair or skewed in the way of product recommendation, pricing strategies
and customer support, all of which could harm the level of trust and overall performance of
the model.

5. Customer experience,

• Uncertain AI powered recommendation.

If the system fails to accurately reflect the customer’s preferences or needs, customers of Daraz
may be prone to feeling uncertain or even dissatisfied with the product recommendations
delivered via AI.
• Pricing algorithms that are opaque
With no transparency in AI driven pricing models, there is a possibility of confusion or
mistrust from customers — especially ones that do not understand the pricing is unfair or
unpredictable.
• Mistrust automated assessments by consumers.
Consumers may be reluctant to grant full confidence to automated assessments or AI driven
decisions like personal offer or review of products that can impact the whole shopping mood
of the experience and their engagement with the platform.

6. Compliance and Ethics

• Regulatory non-compliance risks

Further, Daraz must juggle to comply with local and global regulations such as data
protection laws to avoid the legal and financial risk, while the AI operations themselves need
to comply with local and international regulations too.
• Decision making by AI and ethical issues.
The use of AI when making a decision at Daraz could bring up ethical problems like fairness
and transparency or when an AI system is impacting some critical aspects like pricing,
recommendation, or customer support.
• The difficulty of detecting bias in AI
Bias in AI algorithms is a big challenge for Daraz especially when data or model has a bias
that makes people see something that is not true or unfair and with that happens on customer
trust and the brand reputation.

106 Thanoj E127966

1.7 Stakeholders

A stakeholder is an individual, group, or organization that has a vested interest in the choices
and operations of a company, organization, or intitative. Stakeholders may have no formal
association with the organization in which they have an interest, or they may be members.
Stakeholders may impact an organization’s initiatives or coperations directly or indirectly. The
success of projects and businesses frequently depends on their assistance.
A collection of global guidelines for corporate social responsibility is ISO 26000, which was
created by the international organization for standardization. It provides the following
standards for determining who is stakeholder,
• A company has a legal duty to its stakeholders.

• Decisions made by an organization may have a good or bad effect on them

• They are likely to participate in an organization’s operations and voice concerns.

Based on this criterion, stakeholders generally include consumers, workers, investors,

suppliers, boards of directors, community members and groups, and government bodies.
Stakeholder capitalism is a system in which a firm prioritizes stakeholders’ interests.
Horse racing is where the term “stakeholder” originated. The prize money in a stake race comes
from the money that horse owners pay to enter the event. The entrance fee is referred to ass a
stake, which is another word meaning risk. A stakeholder is the individual or organization that
handles the entrance payments until the prize money is given out. The stakeholder typically
has no financial stake in the race’s result.

Figure 16.Types of Stakeholders

107 Thanoj E127966

Main types of Stakeholders,
Internal and external stakeholders aer the two primary categories into which stakeholders are
frequently devided.
Internal stakeholder,
Those with an interest in a firm that comes from direct employment, ownership, or investment
are known as internal stakeholders. Workers, project managers, boards of directors, funders, a
and investors are examples of internal stakeholders in a business or project. Because they have
a direct stake in the success of the business or project, these people are sometimes referred to
as major stakeholders or key stakeholders.
External stakeholder,
People outside of a firm who are impacted by its choices and results in an indirect way are
known as external stakeholders. Customers, Suppliers, Government, Organizations, Creditors,
Labor unions, and community organizations are examples of external stakeholders. Since their
interest in the business or project is sometimes more symbolic than direct, these organizations
are also known as secondary stakeholders.
(Nick Barney, 2025)

1.7.1 Key Stakeholders in AI for E commerce at Daraz sri lanka

project

Multiple main actors exist within Daraz Sri Lanka to implement their e-commerce platform
with AI technologies which affect operational procedures while determining user interactions
and business operational effectiveness. AI technology integration success at Daraz Sri Lanka
depends on all groups to fulfill their essential functions which secure both operational
excellence and system integration ease. This AI transformation includes multiple stakeholders
who perform the following roles.

1. Customer (Shoppers and Sellers)

Daraz’s (Daraz) AI powered features are being used by customers to get personalized product
suggestions, chatbot assistance and automatic order tracking. With AI they make the shopping
experience more personalized with suggestions of interest, better accuracy in the searches and
the ease of that in the intercourses.
Impact – Higher satisfaction, improved user engagement, and a rise in sales conversions.

108 Thanoj E127966

 2. E – commerce administrators
AI integration is placed under the care of Daraz’s administrative team, with the focus on
inventory management as well as fraud detection across different functions. AI Systems
decisions are brought in line with the goal of business, operating efficiently and adhering to the
industry regulations.
Impact – Better decision making, Increased operation efficiency, and lower risks of fraudulent
transaction.

3. AI Development and Tech Teams

AI engineers, data scientists, software developers and alike are all creating and maintaining the
AI-driven tools such as automated pricing models, demand forecasting algorithms, chatbot
systems etc.
Impact – Accurate data analysis, scalable solutions designed for Daraz’s marketplace, and
ongoing AI advancements.

4. Regulatory Authorities and legal compliance Teams

With AI driving powerful applications at Daraz, the tech giant’s regulatory bodies and in house
legal teams also make sure that the AI applications comply with the laws of consumer
protection, data privacy and ethical AI.
Impact – AI powered businesses follow international data protection requirements (such as the
GDPR) and Sri lankan E – commerce laws.

5. Logistics and Supply chain partners

AI is used to optimize Daraz’s logistics network, automated order fulfillment and AI uses to
predict delivery time. AI powered systems are used by warehouse management, route
optimization and the efficiency of a supply chain.
Impact – Reduced transportation costs, aster delivery, and better inventory control.

6. Marketing and Advertising Teams

It (AI) helps in creating Targeted Ads, based on Customer Behaviour and for Optimising
Marketing Campaigns to name a few. There are tools that use AI to tailor promotions and
automate email marketing all together; they also help in making your customer segmentation
strategy more powerful.
Impact – Higher consumer engagement, improved return on as spent (ROAS).

109 Thanoj E127966

 7. Customer Support Teams
Customer inquiries, order tracking, as well as dispute resolution, are handled by customers
using AI powered chatbots and virtual assistants. AI makes it more responsive and automates
all of the rep’s repetitive or mundane tasks They are doing.
Impact - Higher issue resolution rates, less work for human agents, and improved customer
service.

8. Payment Gateways and Financial Partners

It is important with regard to securing online transactions, detecting the fraud, financial
personalization (e.g., AI powered credit scoring installment payments). Through banks and
fintech companies, Daraz integrates the AI-driven payment solutions.
Impact – Safer transactions, fewer fraud cases, and increased confidence in online payments.

9. Investors and business Stakeholders.

All depends upon how AI-drive decision making works, it is influenced by the investors,
business partners and most importantly, Alibaba because Daraz is the parent company. AI is
exposing business scaling, finances forecasting, and investment ideas.
Impact – Long term business growth, higher profitability, and improved strategic planning.

10. Government and Policymakers

Through policies relating to the digital economy, cybersecurity and taxation of e-commerce,
Sri Lankan government bodies influence AI adoption in e-commerce. They provide the avenue
for employees to balance AI integration benefits to businesses with consumer rights.
Impact – By abiding by the law and moral principle, AI powered e commerce guarantees a fair
marketplace.

As AI is integrated with e-commerce ecosystem of Daraz Sri Lanka, there involved several
stakeholders working together in terms of efficiency, compliance and customer satisfaction. AI
based solutions have the ability to enhance personalization, optimize logistics, fighting fraud
and most importantly better performance of the operations. With the progression of AI, these
stakeholders need to come together to maximize its potential so as to develop the future of e-
commerce in Sri Lanka.

110 Thanoj E127966

1.8 Resource Plan
What is resource planning in project management?
The process of determining projecting, and assigning the most appropriate resources like
personnel, machinery, assests, facilities, and more to projects at the appropriate time and cost
s known as resource planning.
Additionally, it guarantees that resources are used effectively and efficiently throughout the
company. To put it another way, strategic planning is the process of maximizing resource use
based on availability and capacity while completing each project on schedule and within
budget.
Types of Resource planning,
1. Operational planning

2. Strategic planning

Operational planning,
Routine business operations and activities in compliance with organizational polocies, such as
pipeline project management, FTE and contingency resource management, projecting billable
and total usage, and more, are all included in operational resource planning. In order ti achieve
the overarching organizational goals, strategic planning is supported by operational planning,
which deals with a company’s daily operation. We choose thee short term goals first, and then
we figure out how to get there.
Strategic planning,
Top level management typically engages in strategic planning to determine the cource for the
future. It encompasses the whole corporation and is not restricted to tasks inside a certain
department or unit. The long-term resourcing strategy is examined in strategic resource
planning.
It might vary significantly depending on personal preferences. For example, if an IT company
intends to outsource a short-term need, such as improving its IT infrastructure, the resourcing
strategy will go from fulltime employees to contingent workers. Strategic resource planning
will also cover the out-rotation of a specialized, skilled resource from an ongoing project to
launch a new endeavour.
Benefits of Resource Planning in Project Management
• Minimize project resource costs significantly

• Maximize profitable resource utilixation

• Ensures competent resources allocation

• Delivers project on time and within the budget.

111 Thanoj E127966

 Figure 17.Sample Resource Plan

1.8.1 Resource plan for Daraz’s project

1. Human Resources

Role Responsibilities Estimated Cost (LKR) Duration

Project Manager Supervising the budget, 3,500,000 12 months

risk management, and AI
implementation

Data scientists and AI Buiding AI models for 6,000,000 10 months

developers customization and fraud
detection.

E commerce Experts Examining consumer 2,500,000 12 months

data and AI insights

IT and Cybersecurity Protecting AI systems, 3,800,000 12 months

Experts compliance, and data
privacy

Team for Marketing and Implementing AI driven 3,000,000 12 months

Customer Experience ad targeting and chatbots

AI trainers Teaching staff members 2,000,000 6 months

about AI tools
Table 8.Resource plan (Human resource)

112 Thanoj E127966

 2. Technical resources
Resource Purpose Estimated Cost Duration

Cloud Computing AI processing, Data 9,000,000 Ongoing

Infrastructure storage, and model
training
AI programs and permits AI powered 7,500,000 1 time
recommendation and
automation systems

Tools for data analytics Forecasting demand and 5,500,000 1 time

and AI segmenting customers

Cybersecurity Solutions 24/7 assistance with sales 3,200,000 1 time

and customer service

Table 9.Resource plan (Technical resources)

3. Financial Resources
Equipment Purpose (LKR) Duration
Database systems and Managing AI powered 6,500,000 1 time
servers insights and transaction

Design for rapid Networks Endorsing cloud 4,200,000 ongoing

computing driven by AI
AI training Labs AI usage training for 3,000,000 1 time
employees

Table 10.Resource plan (Financial resources)

4. Operational Costs
Activity Purpose Estimated Cost (LKR) Duration
Staff Training and Teaching employees 2,500,000 6 months
Workshops how to incorporate AI

AI Model Testing and Ensuring AI’s 3,800,000 ongoing

Optimization performance and
accuracy.
Legal Recognition and Checking if AI complies 2,000,000 1 time
Compliance with Sri lanks’s
ecommerce regulation
Table 11.Resource plan (Operational costs)

113 Thanoj E127966

 5. Contingency and Maintenance
Category Purpose Estimated Cost (LKR) Duration
Maintenance of AI System optimization and 2,500,000 ongoing
systems routine updates

Troubleshooting in an Resolving security risks 2,000,000 ongoing

emergency and AI faults

Unexpected Expense Filling in unexpected 5,000,000 ongoing

Cushion financial or technical
shortfalls
Table 12.Resource plan (Contingency and Miantenance)

The strategic AI integration at Daraz Sri Lanka’s e-commerce operations properly distributes
all resources—including human and technical elements and financial capacity and operational
assets—to achieve successful AI implementation. Through this complete methodology the
automation of customer experience alongside fraud detection and logistics and marketing
operations simultaneously reaches peak operational efficiency and optimizes complete
organizational processes.
AI based recommendation systems, supply chain estimation platforms and chatbot service
automation helps the organization to reduce their performance flaws and are used to improve
sales effectiveness. Daraz uses AI solutions in multiple operational areas to give accurate
scalable user friendly applications by hiring AI engineers and data scientists in the team and
also employs cybersecurity personnel and support representatives. Security tools, in an effort
to keep customers’ real time, are backed by previous transactions that are made on the cloud
based based systems which make sure of giving customers individualized, effortless shopping.
Financial planning is one of the important steps of AI integration for Daraz Sri Lanka in order
to allocate resources properly. These costs are allocated to AI development, system
maintenance, marketing automation, and to make sure one complies with the regulatory
requirements. Moreover, there is a set aside contingency fund that is dedicated to unforeseen
risks, security threat or scaling challenges so that the AI solutions continue to run smoothly.
Continuous training of the team is a another key focus of the plan and it is because the team
needs to know how to operate, monitor and fine tune the AI models. Daraz considers AI
integration as a long term commitment where frequent performance evaluation, security audits
and periodic update of the AI solutions are done par with the ever changing enterprise goals.
This resource plan will set the stage for the successful execution of which the agency would be
able to pos an as a front runner in the use of AI driven e – commerce, amongst exploiting
innovative technologies in customer engagement, operational efficiency, and a strategic edge
in the Sri Lankan market. With the right AI usage, Daraz will cut costs to the minimum still
resulting in revenue growth, sustain long term growth and finally make it to where it wants to
be in the fast moving eCommerce field.

114 Thanoj E127966

End of Task 1,
From project planning document, the success of AI integration in Daraz Sri Lanka’s e
commerce platform depends on a well structured, phased approach to AI integration. This
project strives to bring significant revolution to Daraz’s operations, proposing way of resource
planning, risk management and quality assurance as well as use of the AI driven solutions to
enhance customer experience, logistics and decisional processes.
Success of integrating AI relies on the project management. Carelessness in any of the phase
can hang a decision or strategy. The objects are defined, stakeholders are discovered, and it is
approved in the initiation phase. In addition, it includes sound technical and financial planning,
risk assessment and budgeting. However, to implement AI we need to build AI models,
integrate systems and human training in order to have smooth implementation of AI. In the
control phase, continuous monitoring, quality assurance, and user acceptance testing is done
for necessary adjustments and refinement. In the final phase, that is closure phase, the created
AI system is reviewed in terms of its acceptability; system transfer is executed; and directions
for future maintenance and scaling are formulated.
Firstly, resource management is a key element and Daraz Sri Lanka has put through a team of
AI and data scientist, IT engineer and business analyst towards the transformation. The
technical resources behind it are the cloud infrastructure, AI powered recommendation engines
and cybersecurity and it is critical to ensure the AI system stays secure and efficient. Financial
planning came in handy as it made sure to cover all the costs, including the costs of the
software, as well as employee training within the budget.
It is critical to use risk management to protect against data security threats, algorithmic bias as
well as integration failures. Continuous testing of AI models, regular updates in cybersecurity,
and staff training programs are the proactive approaches adopted by the project that minimizes
disruptions and risks.
Gantt chart methodology comes handy as it is a structured timeline with set deadlines and
someone accountable for each task and milestone. With this approach, project managers and
stakeholders can look out for such progress, allocate resources optimally, and spot the points
at which things could go wrong before they get rather critical.
It is an important step in the e-commerce operations of Daraz Sri Lanka. AI music consuming
feature such as personalized recommendation An customer service automation And dynamic
price (that will improve customer satisfaction, increase operation efficiency and promote the
expansion of businesses). This AI integration will demand a hefty initial investment but once
up, you stand to gain significantly in revenue, customer retention, better decision making and
such.
In accordance with Daraz’s strategic objective, this project takes place under best practices of
project management, risk mitigation, and quality assurance. The successful launch of the AI
integration project makes Daraz a leading e-commerce player in Sri Lanka that is able to deploy
AI across the e-commerce ecosystem.

115 Thanoj E127966

Task 02,

Research Report

Literature Review
Introduction,
The Sri Lankan e-commerce leader Daraz Sri Lanka consistently adopts technological
innovations to boost operational capabilities and enhance customer satisfaction together with
advancing business expansion. The company continues its innovation path through its
dedication to implement Artificial Intelligence (AI) into its e-commerce systems. The strategic
plan matches worldwide market developments since AI transforms online retail through
personalization services and improved logistics operations and improved fraud prevention and
automated customer service management systems.
Daraz brings the much-needed technology in the form AI driven technologies such as the
recommendation engine, predictive analytics, chatbot automation while utilizing Daraz to
leverage operations to be more efficient and improve customer engagement. Daraz can benefit
from AI powered solutions that will greatly help improve their inventory management, targeted
marketing and dynamic pricing strategy to stay competitive in the very fast changing e-
commerce scene.
However, in spite of these challenges, e-commerce companies have also been or are adopting
or have already been using AI. Moreover, if Daraz executes the implementation up to its
potential with the assistance of AI tech to boost customer experience while reducing operational
expenses, then the future profit is definitely earned. Base on this, the role of AI in e-commerce,
its applications, and benefits, its challenges, and effect on growth of Daraz Sri Lanka in the
future are shown in this literature review.

AI in E – Commerce
With the e-commerce business passing through a unique and changing e-commerce business
necessities period, Artificial Intelligence (AI) is a new venture in e-commerce which is
presently revolutionizing the business by driving complex demanding tasks and expanding the
consumer background and diminishing the work offer. Computing devices which are
simulation of intellectual human capability can perform such tasks as information examination,
design cognition and settling. As these are AI technologies chiefly the Machine Learning (ML)
and Deep Learning (DL) have been a major driving force to the dynamic of these businesses
handling the customers, stocks and more.
Machine Learning is one of the major factors of Artificial Intelligence, which essentially means
making algorithms to work with very large sets of data, find patterns in it, and predict the
outcome without explicit programming. In e-commerce, ML models can predict customer’s
preferences and input that information to predict what a customer will like to buy next. In
addition, these models also help optimize pricing strategy under market trends, competitors
and customer demand for companies to remain competitive. Furthermore, ML is being applied

116 Thanoj E127966

to automated inventory management and demand forecasting of supply chain operations to
streamline the operations, maintaining optimal stock levels and reducing operational cost by
retailers.
Machine Learning is the one of the subset of Deep Learning based on the use of neural networks
to train on with complex unstructured data like image (image dataset), video (video dataset) or
customer review. Now they are using it to make e-commerce platform even more powerful by
giving customers perfect personal suggestion, as engaging and as saley as possible. DL is also
very useful in visual search functionalities to the customers who can upload images of products
to the system and the system will recognize and suggest the available products the they can buy
for purchase. Not only does this visual search enhance the shopping experience but it also
boosts conversion rate of the shoppers owing to an enhanced understanding of shopping.
Furthermore, it’s not only AI used as a personalization nor inventory management tool. In these
days, virtual assistants and chatbots with AI are used to help customers better. These AI
generated tools can talk to the customers in live mode, solving the query, the problem and
giving the directions of how to buy, and the whole journey of the customer from beginning to
the end goes down. In addition, AI is employed for the sophisticated algorithms to uncover
patterns of transactions to spot out of box fraudulent activities and ensure that the security is
secured for both consumers and businesses.
AI helps to revolutionize the e-commerce by providing businesses with a way to offer a more
and faster customized, safe and convenient shopping experience. Machine Learning and Deep
Learning are being applied by companies to analyze vast amount of data in order to boost the
understanding of consumer behavior, optimal operations, and, consequently, to increase
profitability. As AI technology evolves it will have an increasingly pivotal role with regards to
how our future looks when it comes to e-commerce as it will now be able to offer new
possibilities for businesses and customers.

Benefits of using AI in Daraz’s E – commerce industry,

E-commerce is no exception to AI’s transformation of industries around the globe. At Daraz,
AI adoption has been increasingly about improving efficiency, customer experience and
operational effectiveness by magnitudes. The main advantage of AI integration into e-
commerce is to speed up business processes and speed of decisions making. Tasks such as
inventory management, customer support, and marketing, usually go to waste on huge hours
and personnel resources. With the use of AI and automating these processes, AI saves the
valuable time and frees up the resources to work on high paying tasks.AI in e-commerce is one
of the most beneficial things because it can make the ad personalization and accurate. However,
with AI, a lot of customer data can be analyzed and analyzed with a high precision, so these
recommendations and strategies are the right marketing that the customer will buy. Improved
opportunity for a conversion from product suggestion based on customer’s browsing and
buying history contributes to higher sales and customer satisfaction. Similarly, AI also provides
businesses better insight into things like demand trend prediction, stock optimization and
pricing predictions, which in turn means that the customers will be getting the products they
required and will not end up having too much stock and not enough stock.

117 Thanoj E127966

Areas like inventory management, product recommendations as well as customer interaction
through chatbots are easier to automate and make the customers involved at a level while
reducing the cost of operation. By executing automation, product suggestions and interactions
with customers, AI helps enhance the shopping experience and brings extra sales with it besides
providing them with a more personalized experience.
AI also excels in ad personalization. AI is good for Daraz in being able to analyze customer
data with high precision and tap into the level of customer segmentation so to deliver a tailored
marketing strategy and recommendations to individual customers. This kind of personalization,
based on the browser history and purchase history, greatly improves the chances of converting
and the possibility of not only filling sales but also winning customer loyal. Additionally, AI
delivers valuable insights regarding demand forecasting, the stock optimization and price
prediction of Daraz so that it can cater to the demand without oversupplying the market and
without undersupplying the market, improving the overall supply chain of Daraz.
In fact, AI is also an important tool to deliver a better experience – personalized interactions –
for the customers. Daraz’s capability to offer 24 hours a day customer support, reply to queries
and guide buyers without interruption becomes possible due to the increasing demand for AI
based systems. Personalizing the shopping experience.Additionally, the more natural and
customer-centric feeling of interactions increases the loyalty. Security is one of the areas that
an important application of AI is in fraud detection. The AI analysis of the transaction patterns
in the real time alerts Daraz to the unusual activities, saving the customer and the retailer alike
from the frauds. It protects the customer’s sensitive data, as well as building the customers’
trust in the platform.
While using AI to make customer experiences better is only one of many ways it can contribute
to e-commerce operations at Daraz, it also makes it more efficient, and helps facilitate better
decision making. This increased sales and customer satisfaction is made possible by the fact
that AI is able to predict the customer behavior, optimize pricing strategies, as well as offer
products on a personal level. Fraud detection and protection of sensitive information from AI
effectively builds trust and security for customer and operational security. With AI on the verge
of becoming a part of this market integration, Daraz will have a stable e-commerce footing,
optimizing operations, enhancing the customer experience, and ensuring sufficient profit from
the e-commerce.
Potential Risks when Using AI in a company like Daraz,
1. Data security and Privacy issues,
The personal and financial data that Daraz collects is vast, and is necessary for Daraz’s AI
systems to be personalized with recommendations and transactions. But mishandling or breach
sensitive data like credit card details could hurt the company’s reputation, cause financial loss
and destroy customers’ trust. In order to combat these risks, Daraz keeps the data safe using
encryption, access controls, and regular audits to adhere to privacy standard like GDPR for
customer data protection.
2. Bias and Discrimination,
To make such an assumption of historical data as an input to make a prediction is already based
on the assumption that there's no bias on products and if it appears in the history then it would

118 Thanoj E127966

result in either being the prediction or the recommendation. For instance, in such case, if you
train your AI algorithms on the past purchasing data, then you might ignore some qualities in
customers and favor the needs, which will leave them discontent. This bias not only makes the
customer experience bad, it also can act as a legal risk for Daraz. Hence, it is imperative to
have AI models frequently tested and perfected to prevent discrimination and promote fairness.
3. Too much reliance on automation,
Automation is good as it increases the efficiency but there can be consequences of leaning too
much on the AI driven processes and missing opportunities when the human involvement is
needed. For instance, automated chatbots and customer service applications may miss the mark
in addressing sophisticated issues or finding out complete facets of a question posed by end
consumers, which makes it a bad encounter for the consumer. Failure of AI could result in
failure to meet customers' expectations of support, which will result in lost brand loyalty, and
brand trust. For Daraz, it is crucial for them to reach a balance between automation and human
presence in their company, to maintain a good quality of the customer experience.
4. Replacements or Displacing of Jobs because of AI,
AI technologies, like the ones behind laptop customer service, inventory control and targeted
marketing may replace jobs that would be performed manually. Automating the process makes
the job more efficient at the cost of displacing jobs. This transition will be managed by Daraz
when they need to invest in programs for the retraining and upskilling of employees to facilitate
the capacity for adapting to new roles and thriving in an AI driven environment. For the
potential negative impact of AI on employment to be mitigated, while maintaining operational
efficiency under the best conditions, proper workforce management is a necessity.
5. Algorithmic Mistakes,
Thus, mistakes in these systems will mean AI algorithms can recommend the wrong products,
inflate pricing, and cause faulty inventory management among others. For instance, a faulty
pricing can cost your business money and ruin a customer experience, and when you fail to
identify a top selling product you can hurt revenues and customer satisfaction. However, if not
detected and fixed in a timely manner, these problems can quickly develop and lead to Daraz’s
reputation being severely impaired. Having robust monitoring and testing in place to catch and
remedy algorithmic mistakes is a must before that algorithmic mistake matters to customers
and the business.
6. High costs of Implementation,
The upfront costs of AI integration include purchasing and installing AI software, training
employees, and keeping the machine up to date running every process. When a market is very
competitive, the investment has to be made from a perspective of a clear return on investment
(ROI). The risk of financial loss will potentially be immense without a well laid out strategy
and assurance from the technology will work satisfactorily. This means that proper thinking
about the costs and possible long term benefits associated with the adoption of AI has to be
taken into consideration.

119 Thanoj E127966

 7. Moral (Ethical) Issues,
A case for misgivings in ethical practices for adding AI to existing systems is when it comes to
the acquisition, usage and protection of customer data. Daraz knows the importance of
transparency for companies like Daraz. They must be informed about what they are using their
data for and if there is any possibility it will be used in an undesignated way. If this is not done,
they might face public backlash, mistrust and damage to the brand’s reputation. Making sure
that the customer privacy is respected, Daraz must ensure that it should be ethical in data
handling to minimize exploitation and ensure customer confidence.
8. Reliance on technology suppliers,
This can be highly risky, because Daraz will have to rely on third party AI service providers.
When the company is highly reliant on external providers for the AI-powered customer care or
personalized recommendations, any outage of service, data breach or a discontinuation of
service could cause significant operational disruption. This could have major negative impact
on customer experience, business continuity, and even company’s reputation. To prevent such
risks, Daraz must have zonal plans ready and maybe even diversify the technology partnership.
9. Problems with customer Trust,
If a customer feels the interaction with the AI system is not personal and they don’t feel like
they are connecting with the person on the other side, then they are going to hesitate to engage
with it. Customer alienation happens when AI-driven customer support changes from human
to machine contact. What is important for Daraz to remain trusted is to take care of this balance
between AI automation and engagement with human being so that even when a machine for
instance might take over a task in a customer service setting, there should remain the possibility
of personalized assistance from a human being as required, and the feeling of being valued and
understood.

AI Security in Daraz’s E – Commerce operations,

When artificial intelligence is integrated seamlessly into Daraz’s e-commerce webinar, it is
critical that the AI security is standardized to minimize possible risks and protect the platform.
The AI powered fraud detection systems look at enormous quantity of the transactional data in
real-time to identify the suspicious patterns and anomalies associated with payment fraud, fake
reviews or take over of account. AI’s capacity to process enormous quantity of data is helping
Daraz to detect and prevent fraud activities and reducing financial loss and restore trust with
customer.
Fraud detection from the machine learning models are always adapting to new threats without
any manual intervention to improve the fraud detection rate accuracy. AI is also helping with
the aspect of cybersecurity by predicting and identifying weaknesses in the platform and
automating their responses to maintain the trend in cyber or avoid data breaches.
AI is playing a big role in modifying e-commerce experience by making commerce more
personalized and dynamic; as companies are now able to show revenue upcycling as well as
price upcycling in its digital marketing strategies by implementing dynamic pricing, targeted
advertising, and product suggestions among consumers, which can enhance customer

120 Thanoj E127966

engagement and loyalty. Also, virtual assistant that is AI based powered provides the real time
support for the customers to achieve more satisfaction. On the other hand, companies such as
Daraz need to understand that integrating the AI and data must be done ethically to avoid
corrupting customers but also stay honest and confident in customers.
A hybrid approach identifies a hybrid approach, using AI automation for all decisions that are
not security critical but human oversight for all security critical decisions. As continuous
improvement takes place in the development of AI models, Daraz can implement ethical
practices in security of secure AI models, and the integrity of customer data and operational.

AI and Data Management in Daraz’s E – Commerce

The data management at such a fast changing world becomes very important and with the help
of AI integration Daraz can manage such large scale of data more efficently than ever before.
Daraz uses all of that data and AI allows to automate data collection, data organization and data
analysis so that the platform can process and get real time analysis of petabytes of data. By
having this capability, Daraz is able to identify purchasing trends, customer preferences, stock
demand and product offerings, dynamic pricing and supply chain coordination. Furthermore,
Natural Language Processing (NLP) enables AI to retrieve solid consumer sentiment from
client audits that can be utilized to make real-time business decisions.
AI driven data management is better known although it comes with various challenges.
However, the accuracy of data and data quality are very important, as AI systems take into
structured data as well as unstructured data. Business decisions based on inaccurate, biased or
incomplete data, poor customer insight, poor inventory or sales prediction etc will be
inaccurate. Moreover, there is also a question mark coined around data privacy and security
while dealing with information concerning payment details or browsing history. Daraz must
stick to the data protection laws, strict as it is like GDPR, and prevent that personal data be
employed, misused or used without your permission.
To tackle these challenges, Daraz needs to erect strong AI based data governance systems,
monitor and audit the AI models on regular basis, and have transparent policy towards data
collection. With greater encryption capabilities and good ethical data practices, Daraz can
overcome risks and remain a facilitator for the benefits of AI. These data driven insights allow
for better operational efficiency, better customer satisfaction and increase in trust on the
artificial intelligence driven e commerce marketplace.

Theories and Frameworks for AI in Darz’s E – Commerce

On the other hand, there are a number of the key theories regarding the utilization of Artificial
Intelligence (AI) in e-commerce to explain how the AI takes place, how effective the
implementation of the AI in e-commerce is, and how and when the adoption of the AI in e-
commerce is happening. By structuring the understanding of how AI impacts business process,
decision making and consumer behavior in online marketplaces like Daraz, these theories are
helpful. AI will empower Daraz to improve the customer experience, optimize the inventory
management, personalize the recommendations and capabilities available to help Daraz and
the business to adapt to the changing consumer preferences and market trends. Daraz often
121 Thanoj E127966
found itself in a position to compete in the rapidly emerging e-commerce environment and
operate in the most structured way as the impact of AI on both decision making processes and
consumer behavior was becoming crucial.
1. Machine Learning theory,
Machine learning (ML) is one of the key concepts of AI that doesn’t require explaining or
programming the system to learn from data and find patterns based on which it makes
predictions. AI enabled recommendation systems, fraud detection systems, dynamic pricing
strategies are few of those areas in which ML is used in e-commerce. According to the theory,
as long as there is enough data, the AI models will become more and more efficient and accurate
in this area of Ya customer personalization operations over time.
2. Computational Intelligence Theory
This theory uses components of neural networks, fuzzy logic and evolutionary computing to
behave in human manner, on making decisions on complicated surroundings. Computational
intelligence is used in chatbots, AI based automated customer service and personalized
marketing in e-commerce to respond to the queries of an ordinary consumer in a way that is
natural and human like. At the same time, it helps in setting up product placement, as well as
inventory management, prediction of the market trends depending on consumer behavior
patterns.
3. The technology acceptance model (TAM)
Customers’ interactions with and acceptance of AI driven e-commerce platforms are explained
by the Technology Acceptance Model (TAM). This hypothesis states that the two main
elements determining.
Adoption of AI is,
• Perceived Usefulness (PU) – The degree to which AI enhances customer satisfaction
and productivity, as demonstrated via search engines driven AI, tailored suggestions,
and automated support.

• Perceived Ease of Use (PEOU) – How easily companies and individuals can embrace
AI powered features. Customers are more likely to embrace AI powered chatbots and
recommendation systems if thay are simple to use.

For e-commerce companies like Daraz, Amazon, AliExpress, making AI technologies easy to
use and genuinely beneficial boosts customer confidence and engagement.
4. The diffusion of innovation (DOI) theory
The Theory of Diffusion of Innovation (DOI) proposed by Everett Rogers can illustrate how
the matter of AI technology is spread in the e-commerce industry. According to the theory, this
includes ‘innovators’, ‘early adopters’, ‘early majority’, ‘late majority’ and ‘laggards’ who may
or may possibly accept new AI technologies. Businesses which are using AI, namely using AI
for detecting fraud, dynamic pricing, and automated inventory management are first adopters

122 Thanoj E127966

in the field of AI, so to gain a competitive edge. Even so, late adopters will have a hard time
keeping the customers and being operationally sound.
5. Decision Theory and AI
Decision theory used in e-commerce is on the influence of AI driven decision making and it is
the selection of the best possible outcome given the data available. Decision Trees, Bayesian
Networks and Reinforcement Learning models are all used to automate very complex business
decisions by the use of AI algorithms. AI in Daraz enables optimization of advertising
campaigns, pricing of products and forecasting of inventory, leveraging real time data, to
predict the future trends.

Machine Learning Theory, Computational Intelligence Theory, TAM, DOI, and Decision
Theory provide the theoretical framework for understanding the effect of AI on the e-
commerce. These theories explain how AI works, how a user could adopt AI powered solution,
and how AI helps in decision making for different businesses like Dareaz. Aware of these
theories, companies can use AI to create value and add differentiators in the digital market,
innovate, optimize customer experience, and to remain competitive.

(Anon., 1988)
By implementing the key AI theories and frameworks, Daraz can achieve significant
improvement in various areas of its e-commerce operations and at intersecting with real world
industry trends and keep itself competitive. The following is how these theories could be
applied to Daraz’s context:
That said, Daraz leverages machine learning (ML) in order to improve its AI driven
recommendation systems, as well as algorithmic personalization and automation of fraud
detection. The ML models are processing large amounts of customer and transaction data, and
they will continually refine their predictions and recommendations, becoming more and more
accurate as time goes by. In this way, AI recommends how to create personalized product based
on previous browsing and purchasing behavior to increase customer satisfaction and sales
conversion rate. This will further protect both customers and the platform from financial loss
as they will be able to detect fraud in real time determining any transaction patterns. Daraz will
continue to remain competitive in the fast moving e-commerce sector because its platform will
get smarter and smarter to the consumer’s preffernces and trends, hearing and responding.
Daraz can leverage computational intelligence that incorporates neural network, fuzzy logic
and evolutionary computing to boost customer interactions as well as decision-making
processes. For instance, AI powered chatbots can serve in responding in real time, in a human
like way about the customer queries. In possible future, these AI driven companions can rapidly
solve the issue, increase customer experience and reduce the necessity of human touch in the
matter. AI also helps with inventory with purchasing trends analysis and stock level prediction,
popular products are always stocked and inventory is optimised. With continued improvement
of AI, Daraz will be enable to provide a more efficient and personalized shopping experience
that fosters customer loyalty and lowering the operational costs.

123 Thanoj E127966

Technology acceptance model (TAM): It is important to understand the easy of customers and
employees using and seem AI powered feature to be well adopted AI. One thing that Daraz
needs to make sure of is its artificial intelligence systems, for example its chat bots, personal
recommendations, automated customer support etc are as intuitive and user friendly. AI won’t
really earn customer adoption until it’s an inseparable part of the platform and it’s perceived
by customers to improve customer satisfaction (e.g. better recommendations, quicker service,
etc.). To be implemented, the AI at Daraz must be simple and transparent to the point that
customers could benefit from it while they have faith in the technology itself. So with the right
approach, Daraz can actually boost its customer retention rate by providing a convenient and
satisfying shopping experience designed using AI.
Diffusion of Innovation (DOI) Theory: DOI theory is that there will be adoption of AI in e-
commerce in stages, with the innovators and early adopters as the innovators, and then the early
majority, late majority, and the laggards. For an e-commerce firm trying to become a leader,
Daraz, as an early adopter of AI technologies in Sri Lanka, has every opportunity to take on
that role. This way, the company can have a competitive advantage by executing AI for
dynamic pricing, personalization of marketing and automation of inventory management
before competing companies can. Currently, as AI adoption expands, Daraz will have to use its
continuous innovation to keep ahead of the game. It was believed that early adoption will
position Daraz to achieve the brand loyalty and keep the customer satisfied by providing the
latest features which are expected to attract the tech savvy customers.
Decision Theory and AI: For Daraz’s operational efficiency, decision theory is necessary and
being aided by AI. As Daraz uses AI, it can optimise pricing strategies based on consumers’
behaviour, competitor prices, and the market demand. Also, the platform can adjust prices
dynamically to maximize sales, improve margin and keep the prices competitive. Likewise, AI
based algorithms can predict inventory needs by using trends and previous information, to
maintain impressive stock amounts along with consumers wants and needs. Daraz will use
decision theory models like decision trees and Bayesian networks that help them make data
driven decisions faster, so as to reduce human error and generally improve the business
outcomes. Therefore, by automating complex decision in the areas such as marketing, sales,
and customer engagement using AI, Daraz can achieve more consistent, timely, and accurate
outcomes.
Impact of Implementing these theories on Daraz,
Integration of these AI theories to the operations of Daraz would be beneficial to it as it will
bring in so many benefits.
AI will make Daraz do better in terms of delivering a highly personalized shopping experience,
specifically catered to each individual customer’s preferences, behaviour and purchase history.
This will drive repeat purchases, boost loyalty and satisfaction of customers.
Tasks of repetitive nature: AI will do tasks which are repetitive like customer service via
chatbots, inventory management and pricing. The reduction of these operational costs will
reduce the amount of money the company spends to operate the process, and will allow the
company to scale efficiently as it grows.
Early adoption of AI in e commerce: Earliest adoption of AI in e commerce market will put
Daraz in competitive advantage in the Sri Lankan market. Daraz will be able to differentiate

124 Thanoj E127966

itself from other players in the market by implementing some AI driven like features such as
fraud detection, dynamic pricing and personalized recommendations.
Data-Driven Decision: With the help of AI, Daraz can make a well-informed data-driven
decision. This will mean that Daraz is better equipped to adapt to changes in the market with
greater speed and precision, to predict trends, and to optimise inventory and dynamically adjust
pricing.
Better Marketing and Sales: The ability of AI to examine great volumes of data will enable
Daraz to identify customer segments, perfect advertising, and enhance improvement on
advertising crusades. The higher sales would come from personalized product
recommendations; and with the use of automated marketing strategies, the customer
engagement and retention would be enhanced.
But for Daraz to successfully use these theories, these theories, the company would have to
ensure that the data governance, privacy protocols and security measures are up to the mark.
The training and testing of AI models are continuous and must generate the best and ethical
outcomes, and also should remain compliant with data protection regulation like GDPR.

Population of Daraz
The target audience of the study is key departments of Daraz such as the supply chain
management, customer service department and the marketing analytics department, since they
are all key departments to the company’s AI driven e-commerce services. All these departments
are the backbone of Daraz’s digital commerce and logistics ecosystem in which they play a key
role in optimizing inventory management, enhancing customer experience and at the same time
doing data driven marketing which are driven by them. While all three processes aim to provide
better customer service, using AI, supply chain management focuses on streamlining logistics
and warehouse operations, customer service is about the use of the AI-powered chatbots and
support systems, and marketing analytics involves the use of the AI to offer personalized
recommendations and also create tailored advertising. This paper explores how Artificial
Intelligence could be built into such departments and helps in easing operational efficiencies,
increasing customer satisfaction through personalized shopping, and increased data driven
decision making.
It tries to highlight how Daraz is making and maintain its dominant position in the South Asian
e-commerce markets by adopting the latest AI technologies in its processes. In par with Daraz’s
mission of enabling seamless online shopping, the key objective of this study is to explore the
efficacy of AI solutions in automating the supply chain processes, personalizing a customer
experience and improving the fraud detection mechanisms. Within the key operational
departments at Daraz, the research will be conducted to collect qualitative and quantitative data
from logistics managers, data scientists and AI specialists. These insights will be beneficial for
real world application implications and constraints of AI in e-commerce, for making evidence
based recommendations about the utilization and optimisation of AI in digital retail platforms.
Of special importance is also interdepartmental collaboration between logistics and customer
service teams in Daraz as part of this study. Efficient inventory management, real-time
customer support, data driven marketing are all inseparably connected units that need to be
shared in these values, so as to ensure the smooth order fulfillment and personalized shopping

125 Thanoj E127966

experience. Therefore, with the help of these interdepartmental relationships and exploring the
role of AI in e-commerce efficiency, teamwork, and data driven decision making, this project
will highlight the AI in improving Daraz e-commerce efficiency. The findings will ultimately
be useful in the knowledge of AI centricity towards maintaining the modern e-commerce and
provide the foundation for having theories about AI for standardizing AI integration processes
in the industry.

Figure 18. Daraz.lk

Looking specifically at Sri Lanka, Daraz has transformed itself as the leading e-commerce
platform, being a significant part of the country’s digital economy. Based on Daraz Sri Lanka
yearly reports, as of 2022, the number of searches done on the platform reached around 225
million, showing that Sri Lankan consumers are eager to engage with and buy online.
The platform has over 2 million active consumers per month as of which almost 50% are not
based in the Western Province. This is the case of a company that has a really good reach across
diverse parts of Sri Lanka and is being widely adopted, which is something that Daraz also has
a strong presence in terms of adoption.At least daraz.lk received 3.3 million visits in January
2025 regarding website traffic. There are around 67% male and 33% female users at the site
and the bulk of the one's is 25 - 34 years old.
Also, Daraz has been a key motivation in helping the small businesses. There are more than
50,000 small and medium sized (SME) businesses from Sri Lanka on the platform and this
number of businesses increased by 100 percent since the onset of the Covid pandemic.
Overall, Daraz has had a huge and Digital adoption is reached in Sri Lanka and economic
growth driven through its huge user base and strong help for local businesses.
(Anon., 2022)

126 Thanoj E127966

Sampling Methods
It’s rarely feasible to get information from every member of a group while conducting study
on them. You choose a sample instead. The set of people who will actually take part in the
study is known as the sample.
You must carefully consider how you will choose a sample that is representative of the group
as a whole if you want to make reliable inferences from your findings. We refer to this as a
sampling approach. You can employ one of two main categories of sampling techniques in
your study,
• Probability Sampling – By selecting participants at random, probability sampling
enables you to draw robust statistical conclusions about the entire group.

• Non -probabilistic sampling makes it simple to gather data by selecting participants

non-randomly according to convenience or other factor.

In the methods portion of your article or thesis, you should describe in detail how you chose
your sample and how you minimized research bias.
(McCombes, 2019)

4 types of probability sampling methods,

1. Simple random sampling,
The following are essential details about basic random sampling,
• Every member of the population has an qual probability of getting chosen by simple
random sampling.

• Random selection is used. To choose individuals from a population, for instance,

researches could employ instruments such as a random number genarater.

• Bias is cmmon in simple random sampling. A random sample is less likely to be

obtained when the sample size is lower than the population

127 Thanoj E127966

 Figure 19.Simple random sampling

2. Stratified random sampling,

Numerous populations can be separated into non-overlapping groups according to traits that
collectively characterize the community. Generally speaking, this produces more accurate
findings than basic random sampling.
The following are key concepts to comprehend regarding stratified random sampling,
• To guarantee that every subgroup is represented, stratified sampling selects a sample
from each group (or strata) independently.

• Stratifying people based on factors like sex, age, financial level, or ethnicity is frequent.

• Every member of the population should only be assigned to one group, strata must be
exclusive and definite.

• Once a population has been divided into strata, choosen a proportionate number of
people at random from each category. After that, put those people together to from a
sample.

128 Thanoj E127966

 Figure 20.Stratified random sampling

3. Cluster sampling
Cluster sampling divides the population into subgroups or clusters, much like stratified
sampling does. The two probability sampling techniques differ in that regard, though.
With cluster sampling,
• Every cluster ought to share traits with the population. Choose complete clusters at
random rather than picking people from each cluster.

• Make sure the final sample includes each person from each cluster that was chosen.
Choose people at random from each cluster if the clusters are too big.

• Pre-exiting , readily accessible groups are frequently used as clusters by researchers.

Geographical borders, such as cities or counties, serve as the basis for groups, but they
may also be places like offices or schools.
To cut expenses while surveying sizable or widely dispersed populations, researchrers employ
cluster sampling. Nevertheless, these is a greater chance of sampling error with cluster
sampling. It might be challenging to ensure that each cluster accurately represents the whole
population.

129 Thanoj E127966

 Figure 21.Cluster sampling

4. Systematic sampling
Simple random sampling is comparable to systematic sampling, commonly referred to as
interval sampling.
In systematic sampling,
• To create a sample, each person in the population is given a number and then chosen at
predetermined intervals. In other words, each “nth” person in the population becomes
a part of the sample.

• It’s crucial to make sure tha sample frame is free of any hidden patterns that might
influence random selection. The sample may either overrepresent or underrepresent
traits of there is a chance of data tempering.
Because systematic sampling doesn’t require a random number generator and has a clear
selection procedure, it is easier to use than other techniques. However, the selection that results
might not be as random as it would be if a generator had been employed.

130 Thanoj E127966

Advantages of Probability sampling,
• Sampling sizable populations that are representative of your target audience is
economical.

• For populations that are spread out geographically, probability sampling is beneficial.

• Using an agile experience management platform requires minimal technical

knowledge.

Disadvantages of this sampling method,

• Although stratified sampling can provide equal representation of the clusters, it could
not capture all of the variations within the sample population.

• The strata may be divided into a variety of groups using cluster sampling, however
these clusters may share traits.

• Results from simple, random probability sampling can be obtained quickly. However,
the strata and clusters may not be as focused on your target market.

4 types of non-probability sampling,

1. Quota sampling

2. Snowball sampling

3. Convenience sampling

4. Judgmental

131 Thanoj E127966

 Figure 22.non-probabilistic sampling methods

Simple random sampling in Daraz’s project

Steps in simple random sampling,(For instance)
Step 1 (Define the population),
The 10,000 students enrolled in the one university serve as the population in our scenario. The
symbol for the population is N. We may state that our sample frame consists of all 10,000
individuals as we are interested in each and every one of these university pupils. Our sample
frame would be far less than 10,000 students if, for instance, we were only interested in female
university students and excluded all men.

132 Thanoj E127966

Step 2 (Choose your sample size),
Suppose we decide on a sample size of 200 pupils. The number n represents the sample. This
figure was selected because it represents the financial cap and the amount of time we have to
give the students their questionnaires. But we could have also used a sample size calculation,
which is a really helpful statistical tool, to figure out the sample size we nees. This could have
indicated that a bigger sample size, possibly 400 students was required.
Step 3 (List the population),
We must identify each of the university’s 10,000 students in order to access a list of every
student enrolled at the institution, you would have probably needed authorization from student
records (or another university department) if you were truly conducting this research. This is
covered in more detail in the section on the drawbacks of simple random sample.
Step 4 (Assigning numbers to the units),
We now need to assign a sequential number from 1 to N, next each of the pupils. In our
situation, this would involve allocating a consecutive number from 1 to 10,000
(i.e., N = 10,000; the population of students at the institution.).
Step 5 (Find random numbers),
Before choosing 200 students as a sample from the 10.000 students on the list of random
numbers. Either a computer program that genarates these random numbers for you or random
number tables can be used to find these numbers.
Step 6 (Select your sample),
In the end, we decide which of the 10,000 students will be asked to participate in the study.
This would include choosing 200 random number database in this instance. Assume that the
random number table’s first three number were,
001 (the 11th student from the numbered list of 10,000 students)
9292 (the 9,292nd student from the list)
2001 (the 2,001st student from the list)
From out lost, we would choose the eleventh, nine 292nd, two thousandth students to be
included in the sample. We continue doing this until our sample has all 200 of the desired
students.
(Anon., 2025)

With AI powered e commerce eCommerce, data is key to making the needed decisions and
improving user experience as well as business strategy. A key element of Daraz Sri Lanka’s
efficiency is Simple Random Sampling (SRS) to draw unbiased conclusions about customers
and sellers data, as well as assess operational performance. This way of doing things means
that every person in the population gets an equal opportunity to be selected—selecting bias is
taken out of the methodology making it one of the most widespread and effective methods to
learn about AI’s effect on the business of Daraz.

133 Thanoj E127966

Practically, on Daraz research team can examine consumer behavior of a wide range of
customers and the results truly resemble the actual characteristics of the platform’s users.
Typically, the sampling procedure begins with deciding what sample size is proper, and then it
is done by a random process such as game methods and random number generator.
Daraz’s Simple Random Sampling Based AI optimization,
SRS (Simple Random Sampling) is essential for assessing how well Daraz’s AI powered
features work, such as,
• Expert Analysis: The difference between product suggestion created by AI and its effect
on customer interaction and conversion.

• Measuring the efficiency and accuracy of AI chatbots in solving customer queries is

Automated Customer Service.

• Dynamic Pricing Strategies and evaluating its impact of sales and customer satisfaction
in the use of price adjustments with AI.

Experimental AI studies using Simple Random Sampling,

Through the use of SRS, Daraz can also conduct controlled experiments to evaluate the direct
effects of AI driven interventions, such as randomly assigning users to experimental and control
groups. This experimental approach helps measure key performance indicators like,
• These are correlated with repeat purchases: Customer Retention Rates.

• Interestingly, the metrics of interest here include: What contribution do automated

product recommendations have to completing a transaction?

• Evaluating how pricing and promotion driven by AI affects revenue growth sales
performance.

SRS helps Daraz to make data backed optimizations and strategic investments in AI by
ensuring statistical validity. As a method, this provides an objective footing for the analysis of
AI’s participation in personalization, business process automation, and strategic decision
making in the triangle that Daraz Sri Lanka has been built on to fuel the growth of its digital
marketplace AI.

The impact of this sampling method in AI in E – Commerce at Daraz’s research

Data reliability together with decision-making and AI system optimization benefits directly
from Simple Random Sampling (SRS) use in Daraz’s AI-driven e-commerce research.
Research findings from SRS stand as unbiased representations of the complete customer and
seller base when Daraz Sri Lanka utilizes AI to improve user experiences and automate
operations to increase business growth on its platform.

134 Thanoj E127966

With implementation of SRS, Daraz can measure objective Inside of how features of AI like
Dynamic pricing, personalized recommender, chatbot engagement, and fraud detection affect
user engagement and performance of business. The results of the study are because one’s
chance of being selected for the study is equal for every person in the population rather than a
skewed subset of users. By using this method of gathering unbiased data, Daraz can use AI
driven decision making by refining its algorithms based on real customers behavior and not
based on assumptions.
SRS plays another crucial role in supporting the SRS ability to back predictive analytics. This
allows researchers to assess purchasing patterns, seller performance and marketplace trends in
order to assess predicting power in AI models. For example, high quality, randomly sampled
dataset on Sri Lanka’s diverse online shopping demographics is essential for AI based demand
forecasting. It improves stock management, lessens delivery delay and increases customer
satisfaction by effective inventory distribution or distribution.
Furthermore, SRS offers a fair basis for evaluating AI performance, in addition to contribution,
to the testing of AI interventions. Without these personalization, advertising strategies and
automated order fulfilment, Daraz can randomly select its users in control and experimental
groups to find out how AI driven personalization works. This means that statistical significance
becomes a characteristic of AI optimizations in business, which results in the better business
strategies.
Furthermore, SRS helps Daraz’s AI research with minimizing bias in AI training datasets and
reduces the possibility of algorithmic discrimination. If the training data is not representative,
AI systems may have an inherent bias towards certain customers and will unfairly treat them
in recommending or pricing such items. This problem is mitigated by random sampling as it
makes sure that AI models are trained on diversity of user interactions and thus make AI models
more inclusive and equitable.
In the end, SRS also serves to heighten the effect of SRS on Daraz’s AI research due to policy
learning and the benefits brought to data-driven decision making, bias in the AI optimization
process, and improvement of predictive accuracy. SRS can be leveraged by Daraz to refine its
AI systems to enable a much better customer experience, seller success, and to stay ahead in
Sri Lanka’s fast changing e-commerce landscape.

Data types
Data in AI driven e-commerce research can be measured into qualitative or quantitative and
both types of data are equally important in improving the AI model on Daraz Sri Lanka. The
qualitative data (non numerical) like customer reviews, seller feedback and chatbot interactions
are analyzed by the AI systems to provide better recommendations and enhance the customer
service. Such type of data in turn, is useful to understand user sentiment, to find out the issues
with regards to buying experience, and to perfect AI based decision making solutions that are
for instance involved in fraud detection and automation of customer support.
On the other hand, quantitative data is numerical information gathered with the help of several
e-commerce metrics such as volumes of sales, website traffic, rates of conversions, and rates
of click throughs. Structured data is used by AI models in Daraz to detect purchasing trends,
forecast demand and make the best out of price strategies. For example, machine learning

135 Thanoj E127966

algorithms compute dynamic pricing models from browsing behavior and transaction histories
based on which the pricing is done which in turn optimizes revenue with the competitive
pricing. Similarly, inventory management systems that rely on being powered by AI largely
rely on quantitative data to forecast what stock is required, which in turn prevents overstocking
or stockouts.
Gaining from the value of incorporating qualitative and quantitative data types within AI
research at Daraz, AI driven systems become more accurate and efficient at Daraz. Natural
language processing (NLP) algorithms within chatbots and virtual assistants become more
sensitive to a customer’s inquiry with the help of qualitative data. On the other hand,
quantitative data helps Daraz to fine tune the prediction analytics and provide personalized
product recommendations in order to enhance customer engagement and sales. By recognizing
and structuring these data types effectively, AI models can be used to generate insights that are
relevant, reliable and actional and thereby be Daraz’s growth enabler in the Sri Lankan E-
commerce landscape.
Primary Data and Secondary Data,
Primary Data,
Conversely, material gathered by a researcher to solve a particular topic or problem is known
as primary data. It is unique, first-hand, and from an original data source because it hasn’t been
collected yet or might not be available. A researcher uses a range of methods, including focus
groups, questionnaires, interviews, and observations, to gather the data.
For instance, suppose a food manufacturing form was experiencing a decline in market share
and wished to better understand how consumers saw their snanl goods in the context of a
movement in society roward healthier options. One advantage would be that the researchers
couls interview clients using questions that were particularly created to assist them get the
information they needed to solve the business problem. However, as compared to employing
information that is already in the punlic domain, this data collecting strategy could be regarded
as being very expensive and time consuming.
Types of primary data in Daraz’s project,
Primary data collection is one of the key areas on which the AI-driven e-commerce of Daraz
Sri Lanka depends for the improvement of its customers’ interactions and supply chain
operations as well as its organizational decision processes. And this data is from captured
directly from users, transactions and platform interactions and serves as the base for such AI
models as personalized recommendation, fraud detection and demand forecasting. In these
main categories of original information, which are crucial for operations with the e-commerce,
it all depends,
1. Data Transaction,
This includes purchase records, payment details, order history and refund request. For this data,
AI models analyze and find buying patterns, predict future sales, if any and detect fraudulent
activities. The use of transactional data at Daraz enables them to optimize their pricing
strategies as well as to optimize their recommendation algorithms in order to increase
customer’s retention and sales.

136 Thanoj E127966

 2. Consumer Behavior Information,
With such behavioral data (page views, Search queries, click through rates, time taken from
different product page), AI systems of Dreaz accumulate real time data. Using this data, they
give it to AI based recommendation engines so that the AI powered recommendation engine
can display personalised product listings according to user preferences to enhance user
engagement and conversion rates.
3. Data on Seller performance,
AI driven seller management systems need data regarding seller activities like product listing
quality, fulfilment efficiency, customer feedback and return rate etc. This data helps Daraz rank
the sellers, give them insights on their performance and improve the marketplace experience
by highlighting reliable vendors.
4. Data created by users,
The reviews, ratings and queries from the customers are valuable primary data for the AI
models that power sentiment analysis and enhance chatbot interactions. The analysis and detect
of sentiment in Daraz plays an important role to refine customer service responses and prevent
issues in product or services from escalating.
5. Data on logistics and Delivary,
Real-time data became a key deliverable of this thesis; namely, AI systems collect shipping
times, delivery efficiency, last mile logistics data. This is essential for route optimization,
warehouse management and ensuring on time order fulfilment at Daraz Sri Lanka. This data is
used by AI driven supply chain management to minimize the delays and improve customer
satisfaction.

Leveraging these diverse types of primary data, Daraz employs AI to make its operations as
smooth as possible and to prevent fraud while improving the supply chain. With proper
structuring and usage of this data, AI will be able to use the information to make good decisions
that in turn would help the business grow and have operational efficiency in the e-commerce
sector of Sri Lanka.

The importance of Primary Daata in AI for Daraz’s E commerce,

Primary data is critical for improving AI based operations at Daraz Sri Lanka to make more
accurate predictions, provide more personalized customer experience, and follow data driven
business strategies. High quality data, in real time, is a necessity for AI models and in particular
machine learning algorithms to help it identify trends, find patterns and make informed and
right decisions with respect to various functions in a business.
For instance, product recommendation for Daraz uses customer purchase history, browsing
behaviour as well as seller performance data. The platform can hence predict the user
preferences to a high degree of accuracy and deliver a personalized shopping experience,
enhance conversion rates and overall customer satisfaction. Real time transaction analysis also

137 Thanoj E127966

analyses unusual spending patterns, as they help in detecting frauds and making the
marketplace secure.
Within logistics & inventory management, AI takes advantage of primary data such as
historical sales, seasonal trends, live order volumes to minimize inventory issues and
overstocking. Moreover, delivering the greatest amount of goods with the least cost is just one
of the major advantages of the use of AI in logistics: delivery and fulfillment data helps the
logistics solutions be driven by AI to better determine the most efficient routes for delivery,
resulting in faster and more efficient operations.
Daraz’s superior e-commerce establishment in Sri Lanka’s market is attributed to the
continuous refinement of their AI models utilizing accurate and unbiased primary data, all for
customers and businesses that are more gracious.
Primary data gathering AI driven E commerce at Daraz,
Primary data collection is a must for Daraz, which feeds its AI driven operation on the tools of
product recommendations, demand forecasting, fraud detection, and customer experience
improvement. In e-commerce, users, transactions, and platform interactions are the source of
the primary data, they are real time data that’s used as input for business decisions and for
optimization of AI model.
Customers’ interactions on the platform form one of the primary ways that Daraz collects data.
It includes the information from product search, click, time spent on page, cart addition and
purchases. For instance, AI makes use of this behavioral data to study shopping patterns,
understand what items customers often search for as well as what items they tend to purchase,
and then use it to improve its marketing strategies based on that data. Customer feedback and
surveys are also sources of primary data as they help in strengthening customer service and the
seller’s performance, as they include customers’ perception of their satisfaction with the
delivery experience and the quality of the product they purchased.
Another source for primary data collection is the transactional data. When Daraz gets an order
placed, data pertaining to payment, delivery times, and order return and cancellation is
generated. These data are essential for Daraz’s logistics and supply chain efficiency and for
provision of Daraz with AI driven fraud detection and risk management. Additionally, seller
data, such as the product listings, pricing strategies, and response times further allows AI
models to be trained to make facility pricing recommendations as well as identify suspicious,
counterfeit, or poor quality products.
In addition, Daraz cooperates with courier service, as well as its own logistics network, Daraz
Express (DEX), in order to gather logistics and delivery data. The primary data used in building
an AI system is the success rates of the deliveries, estimated times of delivery, real time
tracking, etc. that helps in making a prediction about the delays, optimization of the routes and
a better functioning of the supply chain as a whole.
Additionally, the chatbots as well as customer support interactions make Daraz collect real-
time engagement data. Chat support systems that operate on AI technology peruse the customer
questions, answers to the complaints and the time of service to yield improved automated
responses to the proposed questions and an overall enhanced customer experience.

138 Thanoj E127966

Secondary Data,
According to Salkind (2010), secondary data source is one that already exists and is being
utilized by someone other than the researcher who gathered the original data or for a purpose
for which it was nit intended.
You can modify secondary data to meet your study needs. It can be published summeries or
raw data.large databases of survays, censuses, and social and economic data that are too costly
or impractical for an individual to gather are a few example.
Additional forms of secondary data include market research data, interview or focus group
transcripts, organizational records, and surveys (such as employee surveys). Secondary sata is
a quick and accessible source of information, whether you are gathering material for your own
project ot building an evidence portfolio.
(Anon., 2025)

Types of secondary data in E commerce project,

Secondary data is the information already that has been gathered and analyzed, and put out by
an other source. In a case of the AI-driven e-commerce initiative, Daraz, secondary data is very
important for the market analysis, forecasting the consumer behavior, competitive
benchmarking and business strategy formulation. The basic secondary data kinds of an e
commerce project are as follows:
1. Reports on Market Research,
Statista was deemed to have published reports about industry trends and behaviors of
consumers, and e-commerce tech: from research firm such as eMarketer, McKinsey, and
Statista. This helps Daraz in taking decisions around where to expand, the marketing strategies
to follow and also on the implementation of the AI.
2. Consumer behaviour and Demographic information,
Opportunity exists in being able to look at public data from government agencies, academic
research, and industry publications regarding customer preferences, buying behavior, etc. This
data can be used by Daraz to improve the AI based recommendation algorithms and also to
target the advertising.
3. Reports on Competitive Analysis,
The secondary data is taken from the competitors’ financial reports, pricing strategies and
customer reviews against other e commerce platform benchmarks Daraz. This information is
used by the AI models to find the best pricing of the product, optimize customer retention, and
prov to improve and manage the seller.

139 Thanoj E127966

 4. Data for social media and online reviews,
Information from the Facebook, Instagram, Twitter, and Trustpilot platforms reveals a customer
sentiment, brand reputation and market trends. The secondary data can be processed by AI
powered sentiment analysis tools to improve service quality and hence customer engagement.
5. Scholarly and Business publications,
There are studies about AI, e commerce best practices and case studies of the success of its
deployment in retail published in journals, conference papers and white papers. These insights
can be used by Daraz to optimize its and Artificial Intelligence based on logistics, fraud
detection, recommendation systems.
6. Economic and Government Data,
National statistics offices and trade organizations publish such economic indicators as GDP
growth, rate of internet penetration and online shopping trends. This can be used to help Daraz
understand the potential market, plan for strategic investments and what regulatory changes
that may impact e-commerce for it.
7. Using Open Data Sources and Web Scraping,
Extracted data from public resources, including competitor websites, pricing comparison, and
discussion forums, provide a direction to change Daraz’s pricing, discover trending products,
and transform artificial intelligence based product personalization features.

Gathering Secondary Data in Daraz,

Daraz heavily relies on secondary data collection and use it in support of relying on AI based
decision in things like market analysis, competitor benchmarking, pricing optimization, and
customer behavior insights. Secondary data is different from primary data which is collected
from existing sources rather than directly from users and transactions. They include internal
databases, external reports as well as other public information.
The second source of data is internal business record of Daraz. Past sales report, transaction
history, customers’ purchase history serves as long term trends that can be discerned from
which AI models learn to refine the demand forecasting and personalized recommendations.
Additionally, data from past company’s marketing campaigns such as advertisement’s
performance and customer metrics is used to improve the future advertisement plans.
The other important source of secondary data is external industry reports and market
researches. It uses other market trends, e-commerce research, published in research firms,
government agencies and analysts of the industry to understand consumer spending patterns
and emerging opportunities. In the following time, it will help with business decisions
depending on facts and the development of AI models alongside the changes in the market.
A major portion of secondary data collection is competitor analysis. Web scraping tools help
Daraz obtain competitor pricing, product offerings and promotional campaigns via public
sources on other online marketplaces. This data is used by AI algorithms to change the prices
dynamically such that Daraz continues to be competitive in the market.

140 Thanoj E127966

Second is another source of secondary data, which is customer reviews and social media
insights. AI based sentiment tools scanning customer feedback, ratings, chatter across such
platforms like Facebook, Instagram, or Twitter, try to figure out consumer preferences trends,
hot satisfaction product moments or areas which are spoiling.
Secondly, government statistics, economic reports, whitepapers from industry, amongst others,
are also used in collecting secondary data. By accessing these data sources, Daraz is able to get
updated with changes in laws and regulations, taxation and economics as they relate to online
retail. Additionally, the academic research on AI in e-commerce is also a part in refining
machine learning models so that Daraz stays ahead in the curve of innovation.
Through leveraging these numerous secondary data sources, Daraz benefits from augmented
origin of AI driven analytics, improved operational efficiency, and enhanced capability to
deliver seamless shopping experience to customers.

The distinction between Daraz’s primary and secondary data,

In the realm of Daraz’s AI-driven e-commerce functioning, primary and secondary data have
very special place as it plays an important role for improving business strategies, better
customer experience and efficient platform functionality.
Primary data is the data which has been collected through firsthand interaction of user to
transactions via which technical system also generates some information. It literally covers real
time such as ads seen product search, clicks, time on page, cart addition, purchase. This
categorize includes the transactional data like the time of delivery, payment details and order
cancellation etc. Also, the customer reviews, chatbot conversation and seller performance
analytics serve as the source of additional primary data to refine the AI driven
recommendations, for fraud detection and for better logistics optimization.
On the flip side, the secondary data present here includes already collected data which Daraz
makes use of for the insights. That includes historical records of the internal sales and
marketing campaigns and historical trends in pricing. Other than the primary data, secondary
data comprises of external sources like market research report, competitors analysis,
government statistics, and social media trends etc. Unlike primary information that is exclusive
to Daraz’s platform, secondary information gives a better general industry understanding to
fine tune AI algorithms, forecast market demand and customize pricing strategy.

Possible Difficulties with Data use in Daraz,

Both these types of data are necessary for the AI driven decision making at Daraz but their
collection, accuracy and application are not easy.
1. Data Quality and Accuracy

▪ This is why to be successful with AI supervised decision making at Daraz Sri

Lanka, the data must be of high quality and accurate. Real time data collection
provides some level of efficiency of the platform but errors from human errors,
141 Thanoj E127966
 fraudulent activities or platform malfunctions can affect real time data
collection. If data is not accurate, forecasts will not be either, which makes a
difference in areas like it personalized recommendations, fraud detection, and
logistics management.

In order to do that, Daraz uses very powerful advanced AI models continuously

refine and validate data, filtering out inconsistencies and detecting anomalies in
user’s behavior. Daraz keeps the data integrity in order to make the better
decisions, get better customer experience and improve the business
performance.

▪ In the case of Daraz Sri Lanka, however, depending on secondary data may be
more challenging as such data may no longer be relevant and may not even be
relevant to current market conditions. As e-commerce is a very fast moving
market, real time, relevant data is needed for real time realignments for many
areas such as pricing strategies, demand forecasting, and customer behaviour
analysis, and any real time data that is out of date is unnecessary for the
decisions the AI makes in these areas.

To mitigate this, Daraz collects primary data from the real time user
engagements, transactions and platform analytics and use this on their AI
models only to be able to bring the best possible business potential from them.

2. Compliance and Data Privacy

▪ The first priority of Daraz Sri Lanka while collecting and using customer data
is to comply with GDPR and local data protection laws. When applying AI for
business operations understand that it does not mean to rely only on data in
order to personalize and optimize the business, it must incorporate data to
uphold the privacy of the customers.

This is one of the critical issues when it comes to protecting customer sensitive
data and keeping the customers trust. The data collection, storage and
processing are secured as well as transparent and are done in an explicit user
consent manner. With this, Daraz practices compliance standards while still
allowing the use of AI for e-commerce innovation.

3. Integrating Different Data Sources,

▪ It is important to integrate the hidden truth, the primary data (real time
behaviour) in the real time context, and external insights of the secondary data
(historical data) with the secondary data (historical data) in developing reliable
AI models. Nevertheless, this integration is contingent on robust data processing
frameworks to enable flow of data between multiple sources seamlessly.

142 Thanoj E127966

 Poor integration of primary and secondary data, or poor match between primary
and secondary data, can seriously complicate the productivity level of AI
enables, such as recommendation engines or forecasting demand. To tackle this
challenge, Marchioglio states that Daraz embraces the solutions leveraging
advanced data integration and allows heterogeneous data types to be
synchronized which further allows the AI models to produce accurate and
actionable insights that can play a role in improving customer experience and
optimizing the business decision making.

4. Processing and Storage Expenses,

▪ As a result, at Daraz Sri Lanka data cumulatively has to be handled in large

volumes in real time which inherently demands large amount of computing
resources and storage capacity. As it comes to data collection and process,
advanced infrastructure are required that collect and process data in a
continuous manner especially for AI driven operation such as personalized
recommendation and fraud detection.

While this is a bit expensive, Daraz works towards cloud computing and AI
infrastructure for data storage and processing on a scalable level. The
investments are significant, but they are necessary to keep the highest level of
realtime data analysis to optimize AI operations and to deliver true customer
engagement and optimized business performance.

5. Bias in AI Models,

▪ In particular, at Daraz Sri Lanka, fairness and accuracy in AI models has to be

ensured. Data collected from the primary and secondary sources can cause an
AI system trained on biased data to generate unfair recommendations, pricing
discrepancies or inaccurate demand forecasting. This makes the use of AI
systems ineffectual and can damage the customer experience, poor business
performance, and more.

▪ This is the reason why Daraz is making a big deal on getting diverse and
unbiased data. Daraz makes sure that data that is used while training the AI
models is representing range of customer behaviors, preferences, and market
conditions so that we could preserve the fair, great AI driven solutions that add
value to everyone and have a positive, fair shopping feeling.

6. Competitive analysis and market trends,

▪ So, while at Daraz Sri Lanka, depending on the secondary data from competitors
or market reports alone can be risky since those data may not be accurate or up
to date. This information has become a completely dependent information that
if missed can lead to misaligned business strategies and that would have been
too expensive to miss the opportunities.

143 Thanoj E127966

 For this, Daraz uses secondary data with live data from its own platform
involving customer behavior and transactional data. This planning also helps
Daraz stay competitive, it allows them to quickly assess market trends,
competitive dynamics and keep changing strategies in line with evolving
demands of their customers and the entire e-commerce environment.

Qualitative and Quantitative Data

Qualitative Data,
Qualitative data are those gathered based on category factors. The nature of qualitative data is
more conceptual and descriptive. It evaluates the data according to its category, kind, or
collection.
The sort of quality provided determines the data collected. Qualitative data is divides into many
groups according to certain attributes. The information gathered from these types of studies or
analyses is utilized for perceptions, theorization, and the certain of speculative hypotheses.
Texts, documents, audio and video recordings, transcripts, and other sources are used to gather
these data.
Examples of Qualitative Data,
• Responses in text format to open -ended survey questions

• fieldwork observations or observational notes Transcripts of interviews.

• Images or videos

• Case studies or personal accounts

Types of Qualitative data in Daraz’s project,
E Commerce qualitative data is just non numerical information that is undergone to find out
how fulfilling the stores and customers are to operate. This is unlike quantitative data which
relates to the numbers and numbers in the space, while qualitative data permits one to gather
the data of what customers feel, what they believe, what they are hoping for. On the other hand,
in the case of Daraz and the wider e-commerce industry, there are few types of qualitative data,
1. Customer opinions and input,

▪ Customer reviews and ratings for the product help identify the product quality,
customer satisfaction and common issues.

▪ Daraz’s comments on seller profiles can be used by Daraz to assess how well
seller performs and service quality.

▪ The feedback of customers about the delivery experience outlines the areas of
improvement for the logistical and the supply chain efficiency.

144 Thanoj E127966

2. Conversions with Chatbots and Customer Service,

▪ There are conversations between customers and AI driven chatbots, or human

support agents, that tell you what specific pain points or frequently asked
questions there are along with where the automation can upgrade in relation
with the service.

▪ Customer messages can be used to detect frustration points in customer

experience and to improve customer experience as a whole.

3. Sentiment analysis and social media mentions

▪ Qualitative brand perception and customer expectations can be obtained from

customer discussions, complaints and praise about Daraz on platforms like
Facebook, Instagram and Twitter.

▪ Daraz can read up hashtag trends and what users are talking about and draw
some insight about the future shopping trends and desired product variations.

4. Seller content and product descriptions

▪ Your shopping experience ultimately depends on the way the sellers describe
their products and ways by which they can use the right word and carry out the
right storytelling.

▪ Images, videos, show pieces can allow customers to take informed decisions
while purchasing the products.

5. Patterns of customer journey and navigation

▪ You can quantitative (e.g. times on the pages, bounce rate) navigation data, but
we do not do qualitative. we quantify how users navigate our website / mobile
app.

▪ Customer behavior is helped with learning where customers spend the most
time on a webpage and which aspects of the interface are confusing.

6. Reasons for Returns and Complaints

▪ Qualitative insights about product defects, misleading descriptions or

dissatisfaction in service can be obtained from why customers return them or
complain.

▪ Daraz analyzes complaint message and return requests to assess the quality
control and customer service strategies.

145 Thanoj E127966

 7. Content for Affiliate and influencer marketing

▪ Quantitative data gathered related to reviews, endorsements, and patron’s

purchasing motivation Perception of reward or punishment is based upon
reviews and endorsements from influencers and affiliate marketers.

▪ Influencer campaigns play an important role in creating tone, style and message
of the brand to impact brand trust, consumer decision making.

8. Discussions in forums and communities

▪ Customers use that to share their experiences and opinions in online forums,
discussion groups, and Q&A section of Daraz’s platform.

▪ By engaging in these discussions, Daraz gains a good amount of qualitative

insight into what the customers are worried about and what trends they
recognize in the market.

Importance of Qualitative data in Daraz’s project,

Later on, qualitative data truly works as an important factor in Daraz’s research to uncover
deeper insights of what these customers desire in terms of their buying behavior and shopping
experience. Quantitative data is different from that of qualitative data in the sense that whereas
the former deals with numbers and statistical trends, the latter enables Daraz to understand the
emotions, motivations and challenges customers and sellers go through. Customer reviews,
chatbot conversation and social media feedback can help Daraz gain an insight about pain
points in the shopping journey, and make improvements addressing the shortcomings for better
customer satisfaction. This data is helpful for determining the most common complaints that
customers have, as well as more efficient ways to support the customers, and create the
communication between the buyers and sellers.
Qualitative data helps Daraz tune up its product listings and advertising strategy in marketing
and product promotion. These are the type of customer feedback and product related inquiries
that help unveil the way in which consumers understand the specific items offered, and how
they perceive product descriptions, as well as type of additional information around an item
needed them to feel comfortable to make their purchase. Daraz finds out what’s happening in
consumer trends through influencer content and everything is discussed on social media, so
they can make better marketing campaigns as well as provide recommendations based on
personalization. This also applies to AI powered recommendation engines which, besides the
purchase history, need qualitative feedback at understanding the user interests and thus
delivering a more engaging and relevant shopping experience.
The second critical area in which qualitative data is useful is seller performance. Daraz
monitors customer ratings and reviews to record the seller credibility, product quality and
service efficiency. Negative feedback for misleading product descriptions, poor packaging, or
late deliveries enables Daraz to enforce stricter quality control measures, in order to keep the
reputation of the marketplace at bay. Additionally, qualitative data is essential in logistics and

146 Thanoj E127966

return management as it enables delivery experiences to be improved by understanding
feedback from the customers about the delivery operations. Thereby Daraz understands why
customers do not return items, and why customers return items and it can further work on
improving product quality checks to minimize product dissatisfaction.
Brand reputation also has effects on qualitative data. Inspection of the social media
conversations, customer sentiment, and competitor market movement helps Daraz to anticipate
the divas and handle its brand’s perception problems before they occur. Interviewing
customers, listening to what they say, listening to feedback and making qualitative
improvements that tell the customers that we are listening what they say, we care about their
views, and more than that, we are addressing their concerns. In addition, competitive analysis
through qualitative data helps Daraz strengthen its pricing strategies, enhance the exclusive
deals, and stay ahead in even the e-commerce market.

Quantitative Data,
Quantitative data are those that are gathered based on numerical factors. The nature of
quantitative data is more objective and definitive. It is numerically represented and measures
the values. The basis for gathering data is “how much” oof the amount. In quantitative analysis,
the data is represented as numerical values that may be measured or counted. The information
is taken from matrices, market reports, questionnaires, tests, and other sources.
Examples of Quantitative Data,
• Age, Height, Weight

• Temperature

• Income

• Number of siblings

• GPA

• Test Scores

Types of Quantitative Data in Daraz’s project,

Quantitative data is quantifiable and numerical information for trend analysis, performance
tracking, data driven decision making in e-commerce. Whereas, in contrast to qualitative data
that focuses on subjective insights, quantitative data gives statistical proof of the consumer
behavior, the financial success and the operational effectiveness. Daraz and other e-commerce
markets fit into the following categories on a quantitative basis too.

147 Thanoj E127966

1. Revenue and Sales Information,

▪ An understanding of total sales volume, revenue generated and average order

value show on a profitability indicators.

▪ Conversion rates refer to the rate at which the visitors of your website would
buy the products after looking at them.

▪ Daraz is able to forecast the demand and change the marketing strategies on the
basis of seasonal sales trends.
2. Metrics of Customer Behaviour

▪ The number of website traffic, unique visitors, customers returning help

measure customer engagement.

▪ Bounce rates and cart abandonment rates indicate what is happening in the
shopping experience and the checkout process.

▪ Marketing research from click rates (CTR) on product pages and advertisements
are calculated.

3. Data on Ligistics and the Supply Chain

▪ However, efficiency in terms of delivery times, order fulfillment speed and

shipping success rates are the measurements of the performance of the logistics
operations.

▪ However, efficiency in terms of delivery times, order fulfillment speed and

shipping success rates are the measurements of the performance of the logistics
operations.

▪ The return rates and refund percentages are indications of the product quality or
misleading of the product description.

4. Data on Marketing and Advertising

▪ Daraz’ customer acquisition cost (CAC) indicates how much the company
spends to acquire a new customer.

▪ ROAS (Return on Ad Spent ) is a way to evaluate digital marketing campaigns.

▪ A numerical insight of social media engagement is measured, including likes,

shares, and comments of social media.

148 Thanoj E127966

5. Performance metrics for customer service

▪ Customer inquiries response times, and chat bot efficiency rates are the quality
measure of customer support.

▪ Customer satisfaction scores (CSAT) and Net Promoter Scores (NPS) represents
overall customer experience levels.

▪ A score on at least number of resolved vs. unresolved complaints gauges

customer service effectiveness.

6. Data on Seller performance

▪ Measnurable insights on seller reliability and product quality comes from

ratings and reviews.

▪ Successful transaction rates and order processing speed are the measures of
sellers’ efficiency.

▪ Seller’s reputation and trustworthiness on the platform depends on the

percentage of fulfilled vs canceled orders on the platform.

7. Security metrics and fraud detection

▪ Daraz uses the way of fraudulent transactions and chargeback rate to adjust
fraud prevention system.

▪ The platform detects unusual logins attempts and suspicious account activities
to further secure the platform.

▪ With numerical data, such anomaly detection is done using AI and flags
potential risks in real time.

8. Performance information for websites and mobile apps

▪ It directly impacts on the user experience as regards the page load times and
downtime statistics.

▪ The metrics that can help evaluate customer engagement is app downloads,
active users and session duration.

▪ Data from Heatmap and A/B results for preferred design of website and user
Navigaition available.

149 Thanoj E127966

How Daraz uses Qualitative and Quantitative Approaches,
As a leading e-commerce platform, Daraz makes use of both qualitative and quantitative
research approaches in order to optimize its business operations, improve the customer
experiences and better decision making processes. The impact of these approaches is it gives
the company understanding on consumer behavior, market trends and operational efficiency.
Quantitative methods in Daraz,
Qualitative research is very important at Daraz Sri Lanka because it is used to understand how
the customers perceive, engage and experience their products and services in our platform. It
is a core method of gathering non numerical inputs, the customer feedback analysis. Daraz
collects customer reviews; analyze complaints and observe seller feedback to extract deeper
insight into customer expectations and pains points. For e.g., interactions related to customer
service can be used to gather relevant information about the common problems faced by Daraz
customers and also identify the areas where Daraz can enhance its services.
Daraz also engages in social media listening, staying connected with what’s happening in the
markets and keeping track of what the brand stands for from the social media by following
Facebook, Instagram, Twitter etc. It allows Daraz to understand the public opinion and
experience in the real time.
Additionally, Daraz views its promotional campaigns effectiveness through content analysis,
like product unboxing videos, customer testimonials etc. Daraz then uses heatmaps as a tool to
track responses, along with posting consumer requirements, in order to gain valuable insights
for how users interact with the platform. They use this data to conduct UX research (we are
doing this) to understand how users use the platform and how can Daraz keep optimizing it to
remake the shopping journey better for users and meeting their needs more effectively.
Quantitative methods in Daraz,
Quantitative research plays an immense role in proving and determining the progress and
behaviour of business at Daraz Sri Lanka. First, numerical data is collected to be used in
analyzing a number of aspects of a business operations. Daraz uses tools such as website
analytics e.g., page views, session duration and bounce rate to know exactly how users interact
with the platform. This data will be an interesting market of how a user engages with a website
and how that experience can be optimized to serve the user better.
In addition to user behavior, Daraz closely monitors sales data to analyze customer purchasing
patterns, including order volume, order value, and seasonal trends. By examining this
information, Daraz is able to optimize its pricing and promotional strategies, ensuring it
remains competitive in the market.
In order to collect numerical customer satisfaction data from customers, such as Net Promoter
Scores (NPS), customer service ratings, or other numerical scores, we often use surveys and
structured questionnaires. By helping Daraz measure customer expectations and determining
the areas it needs to improve upon, these metrics alone provide the recipe for improvement.
Further, the evaluation of logistics performance is done through quantitative methods in Daraz.
The company is able to maintain smooth supply chain operations as it tracks the times delivery
takes, the return rates, and order fulfilment efficiency.

150 Thanoj E127966

That is coupled with machine learning and AI driven analytics that help the company a more
efficient processing large data sets to determine parameters such as fraud detection, product
recommendations and customer preferences based on his her behaviour in the past timeframe.
Daraz uses both qualitative and quantitative research methods to combine them and helps to
make the operation of business depend on data. Apart from improving customer satisfaction
and seller performance, this approach contributes to improving marketing strategy optimization
to keep the company abreast of changes in the ever evolving e-commerce landscape.
Suitable data collection method for Daraz’s project,
Primary data collection through Daraz project will fit best because it presents customers' direct
real time interaction and transaction information along with customer’s comment on the
platform. Such data is important for product recommendations, which can be improved through
better customer service, and fraud detection, which can be bettered. By collecting information
from first hand, Daraz can make data driven decisions where the operation and customer can
provide more accurate prediction to users in general.
The term refers to the process of collecting first hand data from customers, transactions and
platform interactions in order to study on the primary level. This is a vital way to collect
information – the right, up to date information, which is only of relevance to the e-commerce
industry and specifically, Daraz’s AI driven operations. Daraz is able to collect primary data
which can in turn be referred to refine their AI models to, among other things, carry out product
recommendations, demand forecasting, detecting fraud and to enhance customer experience.
Daraz collects real time data and its AI always gets optimized by making it continue to optimize
to meet with the changing of needs of customers, providing the most efficient operational
fitment and the best performance of platform.
Daraz uses many primary data collection methods including,
• Relationships with Customers,
Daraz Sri Lanka also gathers the key primary data from the customer's interaction through the
platform to improve the offerings and the shopping experience. Search queries, clicks, time on
product pages and product actions such as adding to cart are examples of these interactions.
This data helps Daraz to understand shopping behavior and product preferences in depth.
This information is necessary to make customer recommendations more personal and to ensure
that our customers are shown products that are consistent with the things they’re interested in
and previously interacted with. Secondly, it has an important part to play in developing targeted
marketing strategies. With these insights, Daraz will be able to enhance customer satisfaction
levels, improve engagement, and harness sales for an improved shopping experience for the
user.
• Data transaction,
Using data from transactions is critical for optimization of platform operations at Daraz Sri
Lanka. Key information like payment methods, how long will an order take, cancellations and
returns are recorded when an order is placed. This transactional data is critical for ensuring that
the business streamlines the following:

151 Thanoj E127966

 ▪ Delivery times: Daraz can also improve the delivery aspect by rationalizing
delivery times and analyzing the data pertaining to how much time is spent on
fulfilling the orders of products.

▪ Payment data for fraud detection: Payment data helps detect and prevent
fraudulent activities in ensuring a safe shopping environment to users.

▪ Cancellations and returns data help Daraz to understand how dissatisfied its
customers were and helps them improve their services.

• Reviews and comments from customers,

With Daraz, Sri Lanka gathers customer reviews and ratings to determine its customers
satisfaction over the products and the seller. This is valuable feedback for us to provide in
regard to which sellers are effective, the overall shopping experience and product quality. Daraz
can use customer’s review and comments to test and improve product and provide better seller
performance and better shopping experience for future customers. As the loop below
highlights, this feedback loop continues so Daraz may adjust based on data to ensure highest
levels of satisfaction for the platform.
• Conversions with chatbots and customer service,
This data is stored by Daraz Sri Lanka regarding customer’s interactions with both AI chatbots
and customer service agents. They give us insights into the customer’s pain points, the
frequently asked questions, as well as the common troubles during the shopping process. Using
this data, Daraz can see how it can improve in areas of customer support and further refine
services to make for a satisfied customer.
• Logistics information,
Since it has its extensive network of delivery, Daraz Sri Lanka collects logistics data like
delivery success rates, estimated delivery time and the real time tracking information. The
delivery routes need to be optimized and the delivery time reduced using this data in order to
achieve an optimized supply chain. This information could add value to Daraz’s logistics
operation, making sure that delivery is always on time and overall customer experience is of
the best quality.
• Seller information,
With reference to the daraz.com.pk.pk website, it collects seller information as product listings,
pricing strategies and the respondents' time taken. The data is used to train AI models that will
advise which pricing strategy will give the best result and which products are fake or poor
quality. Daraz can analyze seller performance and ensure the quality as well as the
competitiveness of the products that are offered in the product listings in the marketplace, in
order to ensure the quality and the possibility of competitive prices for the overall customer
experience and to maintain the standard of all the marketplace.

152 Thanoj E127966

Why Primary data collection selected for Daraz’s project?
Primary data has the most benefits with the Daraz e-commerce project because it enables the
company to publish fresh and firsthand information from its operations and customers, thus,
making it relevant and accurate. For instance, if the business is about product
recommendations, proper data pertaining to a product, audience preferences, appeals, brand,
consumer psychology, competition, etc. is very important for key business parameters (like
product recommendations, logistics and customer experience) to be optimized. Daraz has
access to the most recent information about their customers and the way they act and how they
purchase when they own its primary data. Apart from this, it promotes highly personalized
recommendations as well as marketing strategies that boost customer satisfaction and
conversion rates. In addition, primary data is valuable in enhancing the logistics by
measurements of delivery times, the inventory levels, order fulfillment, creating a smoother
supply chains and doing better customer service. Additionally, primary data allows Daraz to
accurately measure and enhance its business performance according to customer satisfaction
to promotional effectiveness. Using Daraz’s primary data helps train AI models to improve our
features in fraud detection, demand forecasting, and pricing, amongst other things,
1. Relevance and Context,
With direct collection of Customers’ and market primary data, it is very relevant to Daraz’s
specific data needs and business environment. Such data can be collected through primary data
from Daraz’s users, customers and sellers, which helps the company understand consumer
behavior, preference and pain points and in turn help them modify their platform and make
their offerings to cater to particular crowd.
2. Current and Real time Data,
The primary data collection helps Daraz to gather the most recent info in the fast moving globe
of e-commerce. And with the ability to collect real time data into investor’s pockets, Daraz is
always on trend to change its marketing strategy, come up with the most suitable products, as
well as making quick decisions based on the most modern consumption style.
3. Precision and Dependability,
Primary data involves the direct collection of the info from customers and sellers; therefore, it
is more accurate and reliable compared to the secondary data that might simply be out of date
or over generalized. Daraz collects data through surveys, reviews, etc. as such the insights
gained are real by real experiential data and behaviors. Using this data, product
recommendations, inventory management and customer support decisions can be made.
4. Adaptable to Research Requirements,
Primary data can be collected in such a way as to satisfy particular research goals. Say, Daraz
can create surveys or hold interviews to gauge customer satisfaction, find out what are the best
categories the customers are buying, and evaluate the performance of its marketing strategies.
This flexibility satisfies Daraz’s needs for its data collection to be aligned to their business
goals for getting more insights about areas like customer experience or product performance.

153 Thanoj E127966

 5. Improving business plans,
Daraz can validate its business strategies such as how effective personalized recommendation
is or the effect of change in website design through collecting primary data. As we meter all of
the interactions the customers make with the platform, how they see the product; what that they
buy and what is their satisfaction level — it allows Daraz to optimize its algorithms and
ultimately improve that whole shopping experience. In practice, this is very important to refine
our strategies and make sure the strategy resonates with the customer base.

Ways Daraz may enhance AI in it’s e – commerce operations by using primary data
collection.
1. Personalization via Analysis of Customer Behaviour
Way – User feedbacks, web surveys, and tracking website interactions.
Use - Due to the available primary data that has been collected from user search click and
purchase history, the AI models at Daraz can further be improved to come up with better
product recommendation. Once the data has been analysed by the AI it can then be personalise
this shopping experience for the customer and show customers what products are relevant to
their preferences.

2. Optimization of Chatbots and Customer Service,

Way – Customer support chat logs, AI powered chatbot comments, and call center data.
Use - Chatbots at Daraz can also improve their responses using AI, by learning from the real
time alerts sent every time a customer queries or feedbacks. Primary data about interactions
with chatbots and live customer service can be collected, AI-trained to provide better, more
accurate, and more relevant solutions for the customer, meaning improved customer
satisfaction.

3. AI powered inventory Control and Demand Prediction,

Way – Sales information, supplier reviews, and real-time warehouse observations.
Use - Real time sales data of Daraz and purchase trends of its customers can be analysed by AI
at Daraz to predict demand of particular products. In this way, inventory levels will be
optimized, stock outs reduced and supply chain management becomes more efficient.

4. Improvements to AI security and fraud detection,

Way – Transaction information, Fraud reports, and customer complaints.
Use - Daraz’s AI model can employ primary transaction data and fraud report to find such
suspicious activities as fake orders or unfair payments. The system can then be improved to
better detect fraud and overall enhance security of the platform.

154 Thanoj E127966

 5. Pricing strategies assisted by AI,
Way – Market research, rival pricing tracking, and consumer purchase information.
Use - The use of AI can help Daraz to make its own price changes dynamically with respect to
demand, competitor pricing and seasonal trends by making its own primary data from customer
reaction to price. It optimizes revenue and ensures a competitive pricing.

6. Improving AI powered Delivery Optimization and logistics,

Way – GPS tracking information, reviews from customers, and feedback from delivery drivers.
Use - Data being processed in real time in the form of deliveries can be analyzed by AI at Daraz
to help optimize routes, its delivery time as well and improve its efficiency. Using primary data
collected from delivery personnel and customers, AI is able to continuously improve its
logistics models to perform a more superior service.

7. AI training for voice and image search

Way – Image uploads, search feedback, and customer voice inquiries
Use - In order to enhance the value of AI search features like voice and image search, you can
leverage primary data from real customer interaction. AI models can better understand and
match product searches by means of analyzing customer inputs.

Why Quantitative data collection best for Daraz’s project?

Helps in measuring the performance and scalability for the AI applications. Quantitative data
serves extremely well to track metrics which are measurable and check the efficacy of AI
models implemented in e-commerce operations. This enables you to gather hard data in order
to train AI systems, train decision making algorithms, etc.
Why it’s suitable for,
• AI based personalized recommendation engine can be powered using the quantitative
data such as clicks, views, purchase history, cart abandonment rate.

• In this case, AI has the information on sales numbers and the product trends to make
predictions on future demand, which it uses to manage inventory appropriately.

• Consolidated Metrics: Consolidating metrics such as ratings, reviews and sales volumes
is quantifiable and concrete when tracking metrics which could communicate how well
a customer wants driven sales results.

155 Thanoj E127966

 • AI can price optimize: There are quantitative data like price elasticity, competitor’s
prices, buyer’s patterns of buying, etc. that are considered to optimize price.

Methods for Gathering Quantitative Data in Daraz’s project,

Artificial Intelligence in e-commerce can be integrated by effectively collecting and utilizing
quantitative data that will boost decision making, enhance customer experience and help in
perfecting the operations. Numerical data is collected in different methods in order to gather
insights that would Favor an e-commerce platforms like Daraz, this could be done through
various methods of the AI systems. Within this section, I explore five such quantitative data
collection methods that are important for implementing AI in e-commerce.
1. Questionnaires and Surveys,

For multiple-choice and numerical data, surveys and questionnaires are the most common
method of obtaining feedback from customers. What is this data used for — enhancing AI
powered recommendation engine by capturing feedback from customers through these
methods. Survey analyses powered by AI algorithms help determine customer preferences,
buying patterns, and satisfaction. With the help of this data, Daraz can provide personalization
strategies, product recommendations, and precise marketing efforts.

2. Tracking User Behavior and Website Analytics,

Website analytics: This process collects user interaction data, including click-through, bounce,
and browsing metrics data from the e-commerce website. To improve user experience,
customize search results, and maximize product recommendations, AI uses this data. Daraz can
use data on actual behavior in real-time to spot trends and fine-tune marketing while helping
to maximize customer retention.

3. Split testing, or A/B testing,

A/B testing, or split testing, is a controlled experiment that compares two or more variations
of a landing page, product listing, or marketing campaign to see which one performs better.
A/B testing results are used to optimize UI/UX design, pricing strategies and advertisement to
maximize conversion/results with the digital product. Enter AI, which can be used by Daraz
to test multiple homepage layouts and select the one that provides the best user engagement
and conversions.

4. Analysis of Purchase and Transaction Data,

Ecommerce platforms produce enormous quantities of transactional data such as transaction

histories, purchase frequencies, expenditure patterns, etc. AI-powered data analysis allows for
predictive modeling, personalized product recommendations, and fraud detection. At Daraz,
AI can utilize this data to improve demand prediction, optimize inventory management and
design dynamic pricing strategies that drive sales.

156 Thanoj E127966

 5. Predictive analytics and sales forecasting,

Predictive analytics uses historical sales data to forecast future demand, helping businesses
optimize inventory and marketing efforts. AI-powered forecasting at Daraz can analyze
seasonal trends, promotional impacts, and market conditions to make data-driven inventory
decisions. This reduces stock shortages, prevents overstock situations, and ensures efficient
supply chain management.

Analysing Daraz’s AI integration in E commerce,

Despite a growing adoption of e-commerce, customer experience has been minimal as a result
of limited implementation of Artificial Intelligence (AI) – something that can transform current
interplay with customers, including order placement and processing, and largely enhance
business scale. AI solutions, such as product recommendations, personalisation, customer
interactions and inventory can be fully optimised for Daraz the leading e-commerce product
platform in Pakistan. AI system must be user friendly, directly or orderly incorporated with
existing operations, mathematically full of strength, OTC decision making, and scalable to
future development.

The usage of AI is also a key factor for adoption of AI at Daraz. The AI driven features must
be simple and have an intuitive interface for the employees, to be leveraged and used easily by
sellers and customers without much training. Let’s take as an example the AI based
recommendations system, which should be able to offer OpenCV assignment upcoming
regarding product recommendations with just a simple browsing history, purchase history, the
customer preferences. Also, AI enabled chatbots and virtual assistants need to have a client
engagement that is as smooth as human possible, thus reducing the requirement of customer
support while improving user satisfaction. This ensures that the usability prevents a steep
learning curve and would encourage its widespread use in Daraz’s users.

A significant reason for AI deployment is the seamless integration with Daraz’s existing
infrastructure. Daraz’s AI tools should be compatible with the backend systems, order
management process, and logistics network of Daraz and should not create any hindrance to
the ongoing work. Integration based on API will give AI models direct access to real time data
from the inventory databases, marketing platforms and customer service channels, making it
possible and automated, in a smooth manner. AI-powered dynamic pricing models can be
incorporated with Daraz’s sales data to make product prices it but hinges on demand,
competition and seasonal trends for ideal pricing strategies.

Another part to be considered is the analytical power of the AI tools in e-commerce. Daraz’s
AI systems have to process and analyze many huge datasets like customer purchase history,
product performance and market trends so as to use them for action. There are different
machine learning algorithms that can be used in predicting customer behaviour, inventory
stocking and prevention of fraud. AI can analyze the structured and unstructured data sources
such as product reviews, social media sentiment shown by the customers, etc., to extract strong
insights into the customers’ preferences to assist Daraz in improving its marketing and
operational strategies.

The ability to make real time decisions is significant for improvement of the operational
efficiency as well as the customers’ engagement. So they can have these AI powered
recommendation engines, fraud detection systems, automated customer support tools, etc, they

157 Thanoj E127966

will need to take up the data and take action as quickly as they possibly can, and that’s
instanteously. In particular, an AI model can tell who is shopping real time and make product
recommendations to increase mean order value. Similarly, AI based fraud detection systems
would keep record of how transactions are made and whenever there is a suspicious activity,
immediately catch on and reduce the risk of fraudulent orders and chargebacks.

Both scalability & adaptability are important factors for a long run success. With the growth
of the business of Daraz, the number of transactions, customer interactions and product listings
that AI tools need to process at no performance degradation will increase. In addition, to be
able to add to other markets, integrate new technologies and update AI models in line with the
changing demands of their customers, the AI system should be flexible and customizable. For
example, AI powered localization features are able to serve different countries’ shopping
actions and usage of language and culture specific things as Daraz extends to different
countries.

To integrate the powers of AI to the E commerce ecosystem of Daraz, proper deliberation on

the powers usability, system integrations, analytical powers, real time decisions making, scale
ability for example is vital since AI can use Daraz's insights and automation in scaling these
factors that will lead to higher customer experiences, better business operations, and a higher
position in E commerce brands pyramid. If this process of adopting AI technologies were to
kick start, it wouldn’t just improve the efficiency, through AI technologies adoption but will
be data driven decisions, would shape better workflows and will help boost sales performance
and customer satisfaction.

Questionnaire

158 Thanoj E127966

159 Thanoj E127966
160 Thanoj E127966
161 Thanoj E127966
162 Thanoj E127966
163 Thanoj E127966
Feedbacks from Google form,

83.3% strongly agree – This majority group believes that these techniques significantly
improve the relevance of product recommendations, porivdig them with a more precise and
customized shopping experience.

16.7% concur – Although not as strongly, this percentage concurs that the tools effectively
enhance product recommendation.

• All respondents to the study agreed, either strongly or somewhat, that AI driven
recommendation systems increase the relevancy of product recommendations on Daraz,
indicating a resoundingly positive attitude toward these technologies.

33.3% strongly agree – Fewer respondents believe AI technologies greatly improve the quality
and efficacy of customer service.

The majority 66.7% believe that AI applications have a positive impact, despite their agreement
being weaker than that of the first group.

• According to the image, these is broad consensus (Strong or moderate) that AI

applications improve customer service’s speed and quality.

164 Thanoj E127966

60% agree – Just over half of respondents acknowledge that merchandising features driven by
AI make Daraz more enjoyable to shop on.

A smaller group of respondents, however, leave 20 per cent of their answers neutral when asked
how AI will affect the shopping experience.

A strong majority pushes back – A large fraction does not believe AI driven personalization is
a good thing when it comes to shopping.

• The overall picture from the data is that AI driven personalization features positively
impact the Daraz shopping experience, that the majority agree with, some don’t know.

Half agree: Of the respondents, 50% believe that the integration of AI has made the process of
product searches and purchases more efficient.

A large portion is indifferent to the effects of AI on product search and purchase efficiency:
33.3% neutral.

A smaller group strongly believes that AI has greatly improved product search and purchase
efficiency at 16.7%.

• While half of the respondents believe AI has made them more efficient, the majority
are ambivalent and a bit of a minority disagree.

165 Thanoj E127966

A portion of respondents strongly believe that the AI systems with high degree of reliability
can reliably predict the availability based on the customer demand.it’s 33.3%

A majority agree(AI systems will help predict availability based on the demand): 50%

6.5% neutral – Another smaller group is unresponsive to AI's capacity of predicting the
availability of the product based on customer need.

• The data reveals strong belief that AI could predict customer demand to inform
availability and most agreed on the possibility while the minority is neutral about it.

Half respondents are very much in favor that AI driven logistics tools had lowered delays in
delivery services that 50% have strongly agreed.

A significant portion agrees that AI tools have helped save on the delivery time by cutting the
delays to 33.3%.

A smaller group (16.7%) remained indifferent as to the impact of AI on a reduction of delivery

delays.

• How data supports that AI driven logistics tool has had an impact to reduce delivery
delays: a strong consensus, because overwhelmingly those who strongly agree, while
very few are neutral.

166 Thanoj E127966

Half of the respondents agree – 50% agree that Daraz’s AI features address the customer data
privacy and security issues well enough.

A good proportion strongly (33.3%) agree that Daraz’s AI features guarantee privacy and
security of the customer data.

16.7% neutral – A smaller group is indifferent about whether Daraz’s AI features sufficiently
address privacy and security concerns.

• As found in the data, although almost all the respondents have strong belief in Daraz's
AI features to solve their customer data privacy and security concerns, a small part of
them are still neutral.

50% strongly agree – Half of the respondents strongly agree that it is possible to use the AI
such that all the customers on the platform are treated in fair and unbiased manner.

A large portion agrees that AI At the same time, It supports the fairness and impartiality for all
customers 33.3%

16.7% neutral – A smaller group of people are indifferent about AI’s effect on fairness and bias
on the platform.

• Majority strongly agree that AI should manage to provide fair and unbiased treatment
to customers, while percentage who are neutral is smaller.

167 Thanoj E127966

Agreed by 66.7% – The majority of people affirm that Daraz is handling the use of AI in the
e-commerce appropriately.

A smaller portion strongly agrees (16.7%) – A somewhat smaller proportion definitively does
not completely believe that Daraz takes ethical concerns regarding AI in use.

A small group is indifferent as to how Daraz manages the ethical issues concerning AI in e-
commerce (16.7%).

• While most agree, a single voice was heard that is opposed to Daraz’s experience of
ethical concern for AI in e-commerce.

Of the respondents, 66.7% strongly agree – Nearly all respondents agree that the AI apps on
Daraz could do a lot in terms of improving the future customer experience.

Only 33.3% agree – A lower portion may agree that in the future, AI applications could enhance
customer experiences.

• Most of the respondents are very strong to strongly agree that all AI applications on
Daraz have great potential to make future customer experiences.

168 Thanoj E127966

Half of respondents strongly agree – 50% agree strongly that Daraz supplies enough
improvements and updates to its AI systems to satisfy the customers.

A lower number, namely 16.7%, agrees that Daraz uses regular updates of its AI systems to
respond to the needs of customers.

33.3% neutral – There is a very large group that does not think there are suitable updates and
innovations regarding Daraz’s AI systems.

• It says that when it comes to most respondents’ believing that Daraz adjusts its ai
systems to accommodate customer requirements, almost all of them strongly agreed,
while another third is neutral.

169 Thanoj E127966

Activity 3
Future Recommendations

Technical Recommendation,

1. Boost Data management and integration,

In the future, the focus should be on improving data management and integration by setting up
a central data system comprising of the transaction records, data of customer’s behavior and
supply chain. It will help the process of analytics and decision making with the help of better
AI driven analytics. Conducting Sri Lankan e-commerce data governance policies is essential
so data’s authenticity, security and complying of data with Sri Lankan e-commerce regulations
is as accurate and secured as possible to help minimising inaccurate data and breaches.
Moreover, Daraz should also make sure that the AI tools that they are using are seamlessly
integrated with other existing infrastructure of Daraz, including the order management systems,
customer service platforms, and the logistics Network in order to optimize its operational
efficiency. Also, it is important to handle the company’s and customer data securely while
ensuring it is encrypted at rest and unavailable to unauthorized access, providing privacy and
keeping the customer at trust.

2. Improve customer satisfaction via AI powered customization.

Daraz should use AI for customization to make client satisfaction better. Or at least one of the
approaches would be to develop AI based recommendation engines which take into
consideration the browsing history and the past purchases and then present the user with the
most relevant and personalized product recommendations. Furthermore, adding Natural
Language Processing (NLP) into the AI chatbots will improve artificial customer support
immensely by contributing to saving response times and increasing customer satisfaction. In
addition, sentiment analysis in customer feedback can also be utilized to improve the user
interaction with AI systems and resolve customer concerns more effectively, and increase
shopping experience.

3. Boost cybersecurity and AI based fraud detection,

Daraz should also work on developing machine learning models to provide real time detection
of fraudulent activities like unusual purchasing behavior or unauthorized transactions to
enhance its cybersecurity as well as to improve AI based fraud detection. Through using AI-
driven risk assessment tools, pattern analysis can be undertaken to prevent fraud and
chargeback fraud. On one hand, frequent security audits and assessments of potential security
vetures are needed to be sure that AI systems continue to be immune to the prevalent cyber
threats in the case of any infringement. Further to this, anomaly detection techniques can be
used to detect potential security breaches and unauthorized access to the customer accounts,
thus reinforcing the platform’s security features.

4. Boost AI in supply chain management and logistics.

Similar to that, AI based predictive analytics can help Daraz to improve demand prediction
enabling them to better manage the inventory by adjusting stock levels and minimizing stock

170 Thanoj E127966

shortages. AI can also be used to automate the last mile delivery route decision, thus reducing
costs of transportation and increasing speed of delivery. Daraz can now meet better efficiency
goals for fulfillment by combining AI with warehouse automation and robotic process
automation (RPA). Moreover, such quality control systems will also be powered by AI, which
will allow them to inspect the products' integrity in more detail to achieve higher standards of
quality and consequently lower return rates.

5. Continuously Training AI Models and Monitoring Outcomes.

This lets Daraz build a sound AI training framework over continuous cycling of machine
leaning models to stay updated with the latest consumer and market trends. AI algorithms can
then adapt through real time data streams, and by being refined with real time data streams they
can stay accurate to always know what occasion trend is occurring and how to react to it. In
order to monitor AI performance, Daraz can create AI performance tracking dashboards which
will track the major performance indicators such as supply chain outputs, customer retention,
and conversion rates. The product recommendations and the pricing strategies should regularly
be checked by conducting regular audits of the AI decision making process — to prevent biases
and inaccuracies.

6. Integrating New technologies with AI.

Daraz can integrate new technologies like augmented reality (AR) with AI to enhance the
shopping experience by offering features such as virtual try-ons and interactive product
displays. AI-powered voice search can be incorporated to engage customers more effectively
and make the platform more user-friendly. Additionally, using AI in a dynamic pricing model
allows Daraz to adjust product costs based on demand fluctuations and competitor analysis,
ensuring competitive pricing. Furthermore, AI-generated content can be used for personalized
email marketing and automated campaigns through platforms like OmniTroop, helping to
improve customer engagement and drive sales.

Non-technical recommendation,

1. Promote the Open and Ethical use of AI

To prevent discriminatory practices using any kind of AI and promote the open and ethical use,
Daraz would develop clear guidelines, and ethical rules. We communicate transparency, i.e.
informing customers on the usage of the customers’ data for personalization and
recommendations, and how this gives rise to trust. This will allow Daraz to establish an
independent AI ethics committee to indicate the responsible deployment of the AI systems and
strive to follow practically fair and ethics principles regularly. However, this will ensure both
the platform’s integrity and satisfaction of the customer.

2. Training for Employees and change management

Through which Daraz can highlight benefits and the tools that AI would bring over to other
employees for the opportunity of training programs for AI literacy. The key teams involved
into hands-on workshops should be logistics managers, customer support and marketing people
to help them optimize the AI usage for their teams. In addition, changing management will help

171 Thanoj E127966

employees adjust to AI driven automation to make the change to technology as it evolves.
Further, this will help Daraz Sri Lanka to strengthen its core competence as well as position
the capability to operationalize the AI effectively through internal AI expertise.

3. Encourage stakeholder collaboration and involvement

Daraz should foster collaboration between the AI developers, data scientist, and marketing
strategists and logistics specialists to ensure that the goals of the AI objectives align with its
business goals. In order to comply with Sri Lankan laws, and to weather a judgment that may
require migration to other countries with other standards of data protection, engaging with
policymakers, and regulators will be necessary. Besides this, Daraz should also look into
cutting edge AI solutions by either working with digital companies or other parties working in
Aussie AI research space, introducing new solutions to the e commerce sector. Additionally,
this will pave way to develop open communication with customers as to the use of AI in order
to foster trust and keep them abreast on how it can improve the experience.

4. Creation and implementation of policies

To achieve AI implementation success one thing is sure, that Daraz’s need to deviate the AI
deployment from Sri Lankan e-commerce laws as well as Regulations, Consumer Protection,
etc and international AI frameworks. Data privacy, AI accountability and ethical use of AI in
business practices should then be included in institutional policies. Ethical or legal issues
should be prevented on the basis of regulatory compliance by establishing regular checkpoints.
Furthermore, there should be clear rules for AI based pricing models to establish free from fake
pricing discrimination and manipulation for fair and transparent pricing in the market.

5. Promote the creation and acceptance of an AI culture

At Daraz, the company should host seminars and workshops that will cover AI awareness,
sharing of real world case studies of AI implementations and promote the creation and
acceptance of an AI culture. An encouraging culture of experimenting with AI-driven tools
will enable employees to experiment and innovate strategies in the company to boost business
efficiency. Since customers are the most important thing for any organization, incentivizing
employees to succeed with integrating AI solutions into their workflows will enhance
communication with the customers and overall business performance as well. Moreover, by
encouraging interdepartmental cooperation in the fields of customer experience, fraud
detection and logistics, AI based improvements will be promoted throughout the company.

6. Evaluate AI’s impact and scalability

To evaluate AI's impact and scalability at Daraz, periodic reviews should be established to
assess the effectiveness of AI in improving operational efficiency, customer satisfaction, and
financial outcomes. AI solutions should be scaled gradually through pilot programs, allowing
them to be incrementally integrated across various business functions. Advancing AI models
to keep pace with changing market trends, customer behaviors, and emerging technologies is
essential. Additionally, it is crucial to ensure that AI-driven solutions are developed with long-
term sustainability in mind, supporting Daraz’s growth and digital transformation strategy.

172 Thanoj E127966

Justification of whole study,

The evolution of e-commerce in relation to the advancement in Artificial Intelligence (AI)

presents a perfect opportunity for these types of platforms like Daraz Sri Lanka to transform
how the business is run and also how the customer experiences it. Entering the AI tech into e-
commerce is no longer a luxury, but a necessity that ensures the business stays competitive.
This study is justified because the study goes into the possible use of AI to improve efficiency,
profitability and service especially with context of Daraz Sri Lanka, the leading online
marketplace of Sri Lanka.

AI is important for e-commerce due to its power to automate processes, assist in the decision
making and to personalize customer interactions at a scale. For Daraz’s case, of a platform with
thousands of sellers and millions of customers, the deployment of AI tools like
recommendation engines, fraud detection system and chatbots is important. The good thing is
that these tools can simplify your logistics, allow you to optimize the inventory and allow you
to provide a customized shopping experience to your customers. This ensures that Daraz
strategizes and ensures the long term sustainability of the business through understanding that
how these AI driven solutions influence operations and customer satisfaction.

Furthermore, AI implementation at Daraz carries externalities for Sri Lanka’s digital economy.
With this amount of reach, Daraz can set industry benchmark. If Daraz succeeds in
implementing AI in their operations that would serve as a proof of feasibilty and benefits to
other e-commerce platform thereby growing entire sector. Therefore, this study endeavor to
explore and analyze contributions toward this bridging the gap between technological
innovation and its business application in a localized setting.

But they are not all that straightforward when it comes to AI integration. It is important to
understand other issues such as data privacy, implied biases of the algorithms, and high cost of
implementation. Moreover, this study was justified as it seeks to do away with these challenges
and present insights and actionable strategies to overcoming them. The research aims to find
ethical and sustainable deployment of AI technologies through identification of the obstacles
and suggesting the solutions.

In addition, Daraz Sri Lanka’s mission to support the community by enabling commerce is
further aligned with the study. By adding AI driven tools to its service offerings, Daraz will be
able to boost SMEs in the platform so as to help them grow and flourish. It explains how the
study assesses the potential of AI in developing open digital ecosystem by promoting fairness
in competition as well as seller performance.

Secondarily, this research is important for other understanding how and why AI can be used in
e-commerce, especially in emerging markets like Sri Lanka. AI is yet to make an appearance
in developing markets while they have been widely accepted by developed economies. Reading
this study, you receive a unique view of how localized challenges and opportunities lean on the
implementation of AI in e-commerce. This is a roadmap for Daraz and similar platforms to
leverage all the power of their AI while remaining Ethical and Regulated.

Finally, the reasons for undertaking this study involve its multi layered investigation of how
AI could transform e-commerce. Instead of discussing generalities, the research zeroes in on
Daraz Sri Lanka and not only addresses tangible operational gains and hindrances but
additionally stakes a vote on the general subject of AI powered digital transformation in

173 Thanoj E127966

developing economies. This will be an important source of information for politicians,
businesses and academics for the adoption of AI and that engagement in this area occurs in a
way that promotes economic growth, technological advancement and customer satisfaction.

Conclusion of the Project,

Artificial Intelligence (AI) is sweeping the eCommerce industry across the worlds and is
making a lot of business efficient, improving the operations and increasing customer’s
satisfaction more than ever before. AI, which is one of the core drivers of digital disruption
discussed and integrated widely in various industries including those mentioned above, is also
utilized by some of the leading e-commerce platform at Daraz Sri Lanka to evolve the business
ecosystem in the country. Daraz uses AI to develop future Sri Lanka digital commerce through
tools such as recommendation engines, automated customer service systems, fraud detection
systems etc.

Beyond personalisation, customers can expect to get personalised product suggestion based on
what they’ve purchased in the past at Daraz, using AI. Basically, this makes the users engaged
and results in higher conversion rates. In addition, AI maximizes the organisational efficiency
by configuring logistics, forecasting customer demand and improvement of the inventory using
which the products needed be available at timely while preventing from stock out or surplus
inventory. AI based fraud detection systems also assist in securing online transaction for greater
safety to customers and seller. Daraz adds these AI driven tools to the existing mix and
continues to transform Sri Lanka’s e-commerce market and contribute to growth, innovation
and customer satisfaction.

AI driven tools operationalize at Daraz, helping Daraz to automate handful of repetitive task
and allow Daraz employees to do further high added value work. 24/7 customer support during
the day and night, boost the speed of responses and increase user satisfaction. By using
predictive analytics and machine learning algorithms, Daraz can make data driven decisions so
that the company can change dynamically with regard to the market trends and trends in
consumer behavior. These are the capabilities that keep Daraz ahead of its competitors in a
highly evolving digital economy.

While the study holds these advantages, it did mention challenges of integrating AI, like
privacy of data, algorithms biases, and the costs of integrating the same. This is important
because such protection of a customer’s data will help us safeguard and we as well continue to
build up the trust that stakeholders have in us. To tackle these issues, Daraz needs to form
robust governance policies, train staff in the programs, and collaborate with departments to
foster smooth transition of its operations to AI.

From a long-term perspective, the benefits of AI adoption far outweigh the initial challenges
and costs. By doing this right, Daraz can act as a pioneer in the e-commerce sector in Sri Lanka
and create new norms in the industry. AI enabled solutions will definitely improve customer
satisfaction through personalized, secure and reliable shopping experience and will keep your
operational efficient and profitable. This means as AI continues to evolve, Daraz will have
access to new growth initiatives enabled by the revolution that AI will bring about in the realm
of customer service, marketing strategies and logistics.

174 Thanoj E127966

Finally, this study shows the potential of AI in increasing e-commerce at Daraz Sri Lanka by
streamlining operation, personalization and general customer satisfaction. Yet, achieving such
a goal requires tackling the ethical, technological, and organisational challenges with a balance
and strategical take. Daraz can lead Sri Lanka’s e-commerce industry towards a technology
driven future by committing to Innovation and responsible AI practises and pave the way for
sustainable growth as well as digitizing Sri Lanka’s booming e-commerce market by inspiring
confidence among stakeholders and creating a more inclusive society. All these efforts will
keep Daraz as a trailblazer in digital commerce, offering its customers a seamless and
satisfactory shopping experience and helping in the nation's digital transformation.

175 Thanoj E127966

References
Anon., 1988. psycnet. [Online]
Available at: https://psycnet.apa.org/record/1989-97311-001

Anon., 2022. dailymirror. [Online]

Available at: https://www.dailymirror.lk/print/editorialbusiness/Daraz-charting-a-course-for-success-
in-2022/340-228967?utm_source=chatgpt.com

Anon., 2023. business.adobe.. [Online]

Available at: https://business.adobe.com/blog/basics/what-is-work-breakdown-structure

Anon., 2024. geeksforgeeks. [Online]

Available at: https://www.geeksforgeeks.org/what-is-gantt-chart-in-project-management/

Anon., 2024. instituteprojectmanagement. [Online]

Available at: https://instituteprojectmanagement.com/blog/project-life-cycle/#null

Anon., 2025. dissertation.. [Online]

Available at: https://dissertation.laerd.com/simple-random-sampling.php

Anon., 2025. foundationinc. [Online]

Available at: https://foundationinc.co/learn/return-on-investment-roi/

Anon., 2025. https://www.open.edu/. [Online]

Available at: https://www.open.edu/openlearn/money-business/using-data-aid-organisational-
change/content-section-4.1

Kinza Yasar, 2023. techtarget.. [Online]

Available at: https://www.techtarget.com/searchsecurity/definition/risk-analysis

Malsam, W., 2022. projectmanager. [Online]

Available at: https://www.projectmanager.com/blog/quality-management-plan

McCombes, S., 2019. scribbr. [Online]

Available at: https://www.scribbr.com/methodology/sampling-methods/

Nick Barney, 2025. techtarget. [Online]

Available at: https://www.techtarget.com/searchcio/definition/stakeholder

176 Thanoj E127966