Software Delivery At the Speed of AI: Inflectra Invests In AI-Powered Quality

®
1 | 1/20/2024 © Copyright 2024 Inflectra Corporation – Confidential, Prepared for Amazon Web Services
Software Delivery At the Speed of AI:
Inflectra Invests In AI-Powered Quality
®
Presented by
Adam Sandman, May 2024

®
Agenda
▪ Introduction
▪ AI Across the Lifecycle
▪ AI in Inflectra Products
▪ Today
▪ The Future
▪ Questions

®
®
About Me
▪ Adam Sandman was a programmer
from the age of 10 and has been
working in the IT industry for the past
25 years.
▪ Currently Adam is the CEO and Founder
of Inflectra Corporation, where he is
interested in technology, business and
innovation.
▪ Adam lives in Washington, DC, USA
twitter: @adammarksandman
linkedin: https://www.linkedin.com/in/adamsandman/

®
®
Disclaimer: As with any discussion around the
‘state of GenAI’, the capabilities improve daily, so
what is written here may well be superseded in
short order!

®
AI Across the Lifecycle
How AI is Transforming Software Development

®
®
DevOps: The Software Lifecycle

Adding Artificial Intelligence
AI
AI
AI
AI

®
Broad AI/ML Trends for Testing
▪ AI making testing more important than ever. And the
volume of needed testing grows rapidly
▪ Taking into account the nature of GPT we now see that
testing may be required even more than before. GPTs are
dreaming and very unstable (6 fingers, 3 legs etc).
▪ Greater volume - One of the main constraints for the
development of any IT system has always been limited
resources. It is impossible to write more code than the
existing staff of programmers, testers and documentation
specialists can write.

®
▪ But now this factor will be eroded, and, consequently, the
volume and complexity of systems may increase
significantly. Same programmers are going to become N
times more productive. And with them the risks associated
with errors in the system will rise.
▪ And this means that the volume of testing will increase.
We need to do what we did before, but in much larger
volumes.
▪ The responsibility for the result will be more on the
checker than on the programmer (because AI you cannot
punish and hold responsible).

®
▪ The most valuable sources for GPT are Q&A and examples.
It may be used to both train a model and to validate it.
▪ Tests may be considered as Q&A and be both the source of truth
and validation criteria.
▪ The testers, even those without programming skills are becoming
designers, programmers and testers.
▪ Those who can provide Q&A and examples will be those who
develop a system, because the rest will be automated.

®
▪ Is AI a threat for testers? Well, I'd say it is more of a
threat for Hollywood at the moment.
▪ You can generate a movie right from the script (with sound) with
AI. All you need is a script.
▪ You can create a good script from the idea with AI.
▪ But someone still need to check that movie characters have 5
fingers, don't walk naked in the public places etc.
▪ So there are tons of things to be validated in the generated
video. So those who were doing validation are still needed
because those who produce still made mistakes and bloopers.

®
®
We all Remember this Right?

®
General Use Cases for AI/ML in Testing
▪ Machine Learning improving maintenance and reducing
brittleness of traditional UI automated tests
▪ ML/AI models can scan historical testing patterns for
trends and identify future risks
▪ AI can be used to create adaptive security testing tools
▪ ChatGPT was able to automate social engineering
▪ Lack of defined requirements and even interpretability for
AI models and generative AI use cases

®
AI in Inflectra Products: Today
Generative AI in Spira and Rapise

®
®
SpiraPlan®
Agile Program & Portfolio Management
SpiraTeam®
Application Lifecycle Management (ALM)
Introducing the Inflectra® Suite
SpiraTest®
Requirements, Test Management
& Bug Tracking
KronoDesk®
IT Support & Help Desk
Rapise®
Test Automation (Web, GUI, Services)
TaraVault® - Source Code Management
SpiraCapture™
Exploratory Testing

®
16 | 5/28/2024 © Copyright 2006-2022 Inflectra Corporation
®
Enabling AI with SpiraApp™ Platform

®
®
Generative AI Using ChatGPT Plugin

®
®
Generative AI Using Azure OpenAI

®
Live Application Demo

®
®
Code Analysis to Optimize Tests (EotenAI)
Watch a live video:
https://www.inflectracon.com/the-hub/v/k2cljkg3er459ed7djsxf6l9ks464x

®
®
Flaky Test Analysis from Code (EotenAI)
Watch a live video:
https://www.inflectracon.com/the-hub/v/k2cljkg3er459ed7djsxf6l9ks464x

®
®
Code Understanding (EssenceAI)

®
®
“Self-Healing” Tests - Rapise

®
AI in Inflectra Products: Future
AI Features Planned for Spira and Rapise

®
®
Plans in Spira
▪ Instructing the LLM to generate test cases, tasks, and risks
for all the specific types defined in the product template.
Currently, it just asks it to create "tasks," "risks," and "test
cases" generically.
▪ Adding similar functionality onto the test case details
page where it will be able to generate the test steps for a
specific test case automatically
▪ Extending the functionality to the requirements and test
case list pages so you can select multiple items on a grid
and generate the test cases, tasks, etc.

®
®
Plans in Spira
▪ Adding support for generating child requirements/backlog
items directly from the requirements grid and mind-map
views.
▪ Support for additional open source LLM such as Claude,
Llama and Mistral via. AWS Bedrock infrastructure
▪ Analysis of traceability data to generate additional
artifacts (e.g. risks) and analytics using:
▪ Retrieval-Augmented Generation (RAG) to augment LLM model
▪ Vector databases of key metrics derived from Spira data

®
®
Sample Application – Library System

®
®
Rapise: Visual Testing using GPT4-Vision

Software Delivery At the Speed of AI: Inflectra Invests In AI-Powered Quality

Recommended

More Related Content

Similar to Software Delivery At the Speed of AI: Inflectra Invests In AI-Powered Quality (20)

More from Inflectra (20)

Recently uploaded (20)

Software Delivery At the Speed of AI: Inflectra Invests In AI-Powered Quality