A REPORT OF INTERNSHIP

SUBMITTED TO THE SAVITRIBAI PHULE PUNE UNIVERSITY, PUNE

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE THIRD YEAR OF THE DEGREE

BACHELOR OF ENGINEERING
In

COMPUTER ENGINEERING

By

Bajrang Sumant Thakur

Roll No: 305C017
Exam Seat No: T1092304437

Under the guidance of

Prof L. B. Pawar

DEPARTMENT OF COMPUTER ENGINEERING

SINHGAD COLLEGE OF ENGINEERING, PUNE-41

Accredited by NAAC with A+ Grade

2024-25
Sinhgad Technical Education Society,
Sinhgad College of Engineering, Pune-411041
Department of Computer Engineering

Date:

CERTIFICATE

This is to certify that the Internship Report entitled

“AI: Transformative Learning with TechSaksham

– A joint CSR initiative of Microsoft & SAP”
Submitted by

Bajrang Sumant Thakur

Roll No: 305C017

Exam Seat No: T1092304437

is a curriculum-based internship work carried out at TechSaksham, under the guidance

of Prof. L. B Pawar and it is approved by Prof. L. B Pawar for the partial fulfilment
of the requirements of Savitribai Phule Pune University, Pune for the third year of the
degree of Bachelor of Engineering (Computer Engineering) during the year 2024-25.

Prof. L. B Pawar Dr. M. P. Wankhade

Internship Guide Head
Department of Computer Engineering Department of Computer Engineering

Dr. S. D. Lokhande
Principal
Sinhgad College of Engineering
 Internship Report

ACKNOWLEDGEMENT

I would like to express my sincere gratitude to everyone who contributed to the

successful completion of this project, Chatbot using NLP.

Firstly, I extend my heartfelt thanks to my mentors and professors for their invaluable
guidance, constructive feedback, and continuous support throughout the project. Their
insights and encouragement played a crucial role in refining and enhancing the
implementation.

I am also grateful to my peers and colleagues for their support, insightful discussions,
and collaboration, which helped in overcoming challenges during the project
development.

A special thanks to the developers and contributors of open-source tools and libraries
such as Natural Language Toolkit (NLTK), TensorFlow, and other NLP frameworks
that facilitated the implementation of this chatbot.

Lastly, I would like to acknowledge my family and friends for their encouragement and
motivation throughout the project. Their support kept me focused and determined to
achieve my goal.

This project would not have been possible without the collective efforts of all these
individuals.

Student name: Bajrang Sumant Thakur

Roll no.: 305C017

IV TE, COMPUTER(SCOE), PUNE

 ML Developer Department of Computer Engineering
COMPANY DETAILS

• Name of the Company – TECHSAKSHAM

• Address of the Company – A-1105, Arcadia, South City 2, Gurgaon 122018.
• Website Address of the Company- https://techsaksham.org/
• List of the Products/Services/etc. details of the industry – IT Services

• Duration of Internship: 4 weeks

Details of the External Mentor from Company

• Name – P .Raja
• Contact Number- +91 124 408 0107
• Email Address - [email protected]

Company Services

• Provides internships for students in different fields.

• Provide industrial training to students in various domains to nurture them to become Job ready.
• Research and Innovation
• Internship Domains:
1. Applied Artifical Intelligence
2. Stack Web Development.
Key Services: -
• Software Development & IT Consulting
• Cloud Computing & AI Solutions
• Cybersecurity & Digital Transformation
• SEO, Social Media Marketing & Online Advertising.

V
 LIST OF CONTENT

Sr.
Contents Page No.
No.

I. Title Page I
II. Certificate II
III. Completion Certificate III
IV. Acknowledgement IV
V. Company Details V
1. Introduction 1

2. Company Profile 2

3. Work Done & Project Details 4

Project 1: Implementation of Chatbot

using NLP.

4. Tools And Technologies Used 08

5. Challenges Faced & Solutions 09

6. Learnings & Skills Gained 11

7. Conclusion 13

Annexures

Offer Letter
 Chapter-1
INTRODUCTION

With the rapid advancements in Artificial Intelligence (AI) and Natural Language
Processing (NLP), chatbots have emerged as a powerful tool for automating
conversations and enhancing user experiences. Chatbots are AI-driven software programs
designed to simulate human-like interactions through text or voice-based communication.
They are widely used in customer service, virtual assistance, healthcare, education, and
various other domains to provide instant and efficient responses to user queries.

The primary objective of this project, Chatbot using NLP, is to develop an intelligent
chatbot capable of understanding, processing, and responding to natural language inputs
in a meaningful and context-aware manner. Unlike traditional rule-based chatbots that
rely on predefined scripts, this chatbot leverages machine learning and NLP techniques
to analyze text, extract intent, and generate appropriate responses.

NLP plays a crucial role in chatbot development, as it enables machines to interpret

human language with high accuracy. By utilizing NLP libraries such as Natural Language
Toolkit (NLTK) and deep learning frameworks like TensorFlow, this project aims to
create a chatbot that can learn from conversations, improve over time, and provide more
human-like responses.

Key Features of the Chatbot:

Intent Recognition: Identifies user intent using NLP techniques.

Context Awareness: Understands conversation context for better responses.

Machine Learning Integration: Uses deep learning models for response generation.

Multi-Domain Applications: Can be deployed in various fields such as customer

support, healthcare, and education.

The growing need for automation and real-time communication has made chatbots an
integral part of modern digital systems.
 Chapter-2
COMPANY PROFILE

2.1 Company Profile

The TechSaksham program embarked on a transformative journey with great
enthusiasm with Artificial Intelligence (AI) and Cloud Computing. These tracks were
carefully designed to cater to the evolving demands of the tech industry, providing
students with hands-on experience and real-world projects to bolster their
understanding.
The program’s impact was evident as it quickly gained traction, attracting students
from across the nation. Over time, TechSaksham expanded its reach to cover 7 states,
establishing its presence in major educational hubs. The program’s growth was
astounding, with approx. 19,000 students enrolling in its courses, all eager to explore
the world of AI and Cloud Computing.
The program’s impact extended beyond the classroom, leading to over 2200 successful
placements in more than 370 renowned companies. The industry embraced
TechSaksham graduates wholeheartedly, recognizing their practical skills and
readiness to contribute from day one. TechSaksham also fostered strong relationships
with more than 1200+ colleges. These partnerships allowed the program to reach a
broader audience and empower students from various educational institutions,
transcending geographical boundaries.
2.2 Dedication

TECHSAKSHAM demonstrates its dedication is by employing a team of highly skilled

and experienced software developers and designers. These professionals work closely
with clients to understand their unique needs. TECHSAKSHAM underscores its
commitment to excellence by assembling a team of highly skilled and experienced
software developers and designers. These professionals, possessing a wealth of
expertise, collaborate closely with clients to gain a profound understanding of their
distinct needs and challenges. The hallmark of TECHSAKSHAM approach lies in
crafting bespoke solutions that precisely align with the specific requirements of each
client.

The team's dedication extends beyond mere technical proficiency; it encompasses a

genuine commitment to fostering collaborative partnerships. By working closely with
clients throughout the development process, TECHSAKSHAM's professionals ensure
that the solutions not only address immediate needs but also contribute to the long-
term success and growth of the client's business.
2.3 Strategy

TECHSAKSHAM as a software development company, they likely focus on delivering

customized software solutions to their clients that are tailored to their specific needs
and requirements. This would involve working closely with clients to understand their
unique challenges and goals, and developing software solutions that address those
challenges and help them achieve their goals. The company aims to develop software
solutions that not only address these challenges but also contribute to the realization of
their clients'goals.

It appears that TECHSAKSHAM operates as a software development company with a

client-centric approach. Their primary focus likely involves delivering bespoke
software solutions tailored to the individual needsand requirements of their clients.
This process would likely entail close collaboration with clients to gain a
comprehensive understanding of their specific challenges and objectives.

2.4 Work Culture

TechSaksham is known for its dynamic and inclusive work culture, fostering
innovation, collaboration, and continuous learning. The company values its employees
and provides a supportive environment that encourages professional growth and skill
development.
At TechSaksham, open communication and transparency are key aspects of the
organizational culture. Employees are encouraged to share their ideas and insights,
while the leadership remains receptive to feedback, ensuring that every voice is heard.
This collaborative approach allows teams to work efficiently towards common goals,
driving impactful solutions and delivering the best results for clients.

2.5 Dedication

TECHSAKSHAM demonstrates its dedication is by employing a team of highly skilled

and experienced software developers and designers. These professionals work closely
with clients to understand their unique needs. TECHSAKSHAM underscores its
commitment to excellence by assembling a team of highly skilled and experienced
software developers and designers. These professionals, possessing a wealth of
expertise, collaborate closely with clients to gain a profound understanding of their
distinct needs and challenges. The hallmark of TECHSAKSHAM approach lies in
crafting bespoke solutions that precisely align with the specific requirements of each
client.
 Chapter-3
WORK DONE & PROJECT DETAILS

3.1 PROJECT: Implementation of Chatbot using NLP.

Objective

The primary objective of this project is to develop an intelligent chatbot that can
understand and respond to user queries using Natural Language Processing
(NLP). The chatbot aims to:

• Process and interpret human language efficiently.

• Provide meaningful and context-aware responses.

• Enhance user interaction through automation.

• Improve customer service, virtual assistance, and query resolution.

Technology Used

Component Technology/Tool Used

Programming Language Python

NLP Libraries NLTK, spaCy

Machine Learning TensorFlow, Scikit-Learn

Framework Flask (for web integration)

Database SQLite / Firebase (if used for data storage)

Implementation:

• The chatbot was developed using Python and NLP techniques to process
and generate responses. The implementation involved:
1. Data Collection and Preprocessing:
• Gathered training data consisting of user queries and responses.
• Performed text preprocessing, including tokenization, stopword
removal, stemming, and lemmatization.
• Used one-hot encoding for categorical variables to make the data
suitable for machine learning models.
• okenization: Splitting text into words.
• Stopword Removal: Removing unnecessary words like is, the, and.
• Stemming & Lemmatization: Converting words to their base form
(e.g., running → run).
• Vectorization: Converting text into numerical format using TF-IDF
or Word Embeddings.
2. Intent Recognition & Model Training:
• Implemented a rule-based and retrieval-based response mechanism.
• Used sequence-to-sequence models for generating meaningful replies.
• Naïve Bayes Classifier for text classification.
• Deep Learning (LSTM, RNN) for complex text understanding.
• BERT (Bidirectional Encoder Representations from Transformers)
for advanced NLP processing.
3. Response Generation:

• Implemented three response mechanisms:

• Rule-Based Responses (Predefined if-else conditions).

• Retrieval-Based Responses (Similarity matching using NLP models).

• Generative Responses (Using deep learning models like Seq2Seq and

Transformer-based models).
Key Steps in Implementation:

• Understanding the Problem: Define chatbot use cases and user needs.
• Data Preparation: Collect and preprocess chatbot training data.
• Building the NLP Model: Train intent recognition and response
generation models.
• Developing Chatbot Logic: Implement rule-based, retrieval-based, or
generative approaches.
• Integration & Deployment: Connect chatbot with web, mobile, or
messaging platforms.
• Testing & Optimization: Evaluate chatbot performance and improve
based on real user feedback.

Results:

1. Successful Query Resolution

The chatbot successfully classified user queries with high accuracy.

It provided relevant and context-aware responses.

2. Improved User Experience

Faster automated responses reduced wait times.

Memory retention improved conversation flow.

3. Scalability & Future Enhancements

Voice-based interaction can be added for better accessibility.

Multilingual support can be implemented for diverse users.
AI-driven self-learning can improve chatbot performance over time.
 Chapter-4
TOOLS & TECHNOLOGIES USED

• NLTK (Natural Language Toolkit): – Used for tokenization, stemming,

lemmatization, and stopword removal.
Helps in processing human language data efficiently.
• spaCy: – Used for dependency parsing, named entity recognition (NER), and
part-of-speech (POS) tagging.
• TextBlob: – Used for sentiment analysis and text classification.
Provides simple API for NLP operations.
• TensorFlow: – Used for building deep learning models for chatbot responses.
Supports LSTM, RNN, and Transformer models for advanced conversation
handling.
• Scikit-Learn: – Used for text classification and intent recognition.
Implements algorithms like Naïve Bayes, SVM, and Decision Trees. Git –
Version control system for managing code and collaboration.
• BERT – Used for understanding context and meaning in conversations.
Helps in improving chatbot accuracy for complex queries.
• Flask: – A lightweight web framework used to build the chatbot’s API.
Enables chatbot integration with web applications.
• Django (Optional): – A Python-based backend framework used for handling
complex applications.
Supports chatbot integration with databases.
• Heroku / AWS / Google Cloud: – Cloud platforms used for deploying the
chatbot online.
Provides scalability and accessibility.
• SQLite: – Lightweight database used for storing user queries and responses.
• Firebase: Used for real-time data storage and retrieval.
• MongoDB (Optional): A NoSQL database used for scalability and high-
performance storage.
 Chapter-5
CHALLENGES FACED & SOLUTIONS

1. Understanding User Intent Accurately

• Challenge:
Users may phrase similar questions in different ways, making it difficult
for the chatbot to understand intent correctly.
Ambiguous or incomplete user queries lead to misclassification of intents.
▪ Solution:
* Used BERT (Bidirectional Encoder Representations from Transformers)
to enhance intent recognition.
* Implemented TF-IDF and word embeddings to improve text
understanding.
* Trained the chatbot with a larger dataset containing variations of queries
to improve accuracy.

2. Handling Context and Long Conversations

• Challenge:
▪ The chatbot struggled to remember previous user inputs, leading to context
loss in long conversations.
▪ Difficulty in maintaining multi-turn dialogue where responses depended on
previous interactions.
▪ Solution:
▪ Used session-based memory management to track ongoing conversations.
▪ Implemented LSTM (Long Short-Term Memory) networks to remember
past interactions.
▪ Stored important chat history in a database to retrieve relevant context when
needed
 3. Generating Relevant and Meaningful Responses

• Challenge:
The chatbot sometimes provided generic, repetitive, or irrelevant
responses.
▪ Hard to generate dynamic answers instead of pre-defined static responses.
▪ Solution:
Implemented hybrid response generation:
▪ Generative Responses using Transformer models for dynamic answers.
* Used Reinforcement Learning to improve chatbot responses based on user
feedback.
▪ Retrieval-Based Responses for recognizing similar queries.

▪ Handling Spelling Mistakes & Slang:

▪ Challenge:
Users often make typos, use abbreviations, or informal language, which the
chatbot could not recognize.
▪ Solution:
▪ Integrated Levenshtein Distance algorithm for spelling correction.
▪ Used word embeddings to detect and handle slang.
▪ Implemented customized regex filters to normalize informal text.
 Chapter-6
LEARNINGS & SKILLS GAINED

Technical Skills:

1. 1.1 Natural Language Processing (NLP)

▪ Gained a deep understanding of text processing, intent recognition, and
response generation.
▪ Worked with NLP libraries like NLTK, spaCy, and TextBlob for text analysis.
▪ Implemented word embeddings (TF-IDF, Word2Vec, BERT) to improve
chatbot accuracy.
2. Machine Learning & Deep Learning:
▪ Applied Supervised Learning (Naïve Bayes, SVM, Decision Trees) for intent
classification. Explored Neural Networks (LSTM, RNN, and Transformer
models) for generating meaningful responses. Researched Reinforcement
Learning techniques to enhance chatbot behavior over time.
3. Database Management:
Worked with SQLite, Firebase, and MongoDB storing chatbot interaction.
Implemented session-based memory management to maintain user context
conversations.
4. Web Application Development:
▪ I strengthened my Django skills, particularly in backend development,
handling file uploads, creating interactive user interfaces, and ensuring
responsive design. This experience directly translates to developing full-stack
applications, which is a highly sought-after skill in web development.
5. Real-Time Data Processing:
▪ In the "Data Pipeline for DFQ Processing" project, I learned how to automate
data retrieval, parsing, and processing in real-time, which is essential for
building systems that require dynamic data updates. This skill will be crucial
in developing scalable applications that handle live data streams.
6. Database Management:
▪ I worked with databases (SQLite) to store and manage tray cleaning data,
developing efficient queries for data retrieval and updates. This strengthened
 my database management skills, which are vital for any backend-heavy
application.
7. Deployment & Debugging:
▪ I deployed multiple Django applications and debugged issues related to
integration, model loading, and real-time updates. This experience taught me
the importance of testing and deployment pipelines, skills necessary for any
full-stack or backend development role.

Soft Skills:

1. Problem-Solving & Critical Thinking:

▪ Tackling technical challenges, such as improving model accuracy, parsing
complex data formats, and designing real-time data systems, enhanced my
ability to break down problems and implement effective solutions.
2. Time Management:
▪ Balancing multiple projects with tight deadlines taught me how to manage my
time effectively, prioritize tasks, and stay organized. This is essential in any
fast-paced work environment or when juggling multiple responsibilities.
3. Teamwork & Collaboration:
▪ Collaborating with mentors and team members during the internship helped
me improve my communication skills and learn from others’ perspectives. It
also reinforced the value of working collaboratively on complex projects to
achieve shared goals.
4. Adaptability:
▪ Encountering various challenges, such as dealing with complex data and
learning new frameworks on the go, helped me develop the ability to quickly
adapt to new tools, technologies, and project requirements.
5. Communication Skills:
▪ Explaining technical concepts to non-technical stakeholders, writing clear
documentation, and presenting project updates helped refine my
communication skills. Being able to translate complex technical ideas into
understandable terms is vital for any technical role.
 Chapter-7
CONCLUSION

1. Successful Implementation: The chatbot using NLP was successfully

developed, demonstrating its ability to understand and respond to user queries
efficiently.

2. Advanced NLP Techniques: The project involved the application of text

processing, intent recognition, and response generation, enhancing the
chatbot’s accuracy and contextual understanding.

3. Machine Learning Integration: Implementing models like LSTM and

Transformers improved the chatbot’s ability to handle conversations more
effectively.

4. API and Database Management: The integration of APIs and databases

enabled seamless communication and efficient storage of user interactions.

5. Cloud Deployment & Scalability: Deploying the chatbot on cloud platforms

like Heroku and AWS ensured better accessibility, performance, and
scalability.

6. Overcoming Challenges: Addressing issues like ambiguous user inputs,

response accuracy, and maintaining conversation context strengthened
problem-solving and debugging skills.

7. Skill Development: The project contributed to hands-on learning in AI,

machine learning, web development, and database management, enhancing
overall technical expertise.

8. Practical Industry Relevance: The chatbot’s integration with platforms like

Telegram and web applications demonstrated its real-world applicability in
various industries.

9. Professional Growth: The experience improved teamwork, research skills,

and the ability to work on complex projects, preparing for future industry
challenges.