Sandeep Garg
#ATAITF: A Festival of AI Testing Tools & Techniques
About Speaker
Sandeep Garg
Principal QA Architect
Bridgetree Research Services
Sandeep is a software testing professional and a lifelong learner across multiple disciplines be it software testing, leadership, self-education or parenting.
Currently serving as a Principal QA Architect at Bridgetree Research Services, Sandeep leads a team focused on ETL, Data Analytics, API testing, and ML/LLM backed solutions while maintaining his hands-on testing and useful test leadership. He practices delivering excellence consistently, building lasting customer trust, and helping everyone around him grow. As artificial intelligence disrupts product development, renovation and testing, Sandeep is enthusiastically moving into Machine Learning, MLOps, and AI-enabled products testing—what he sees as a forward moving and golden opportunity for thoughtful testing professionals.
Away from work, Sandeep recharges by diving into books and researching white papers on diverse topics that impact human lives. He loves spending quality time with his twin daughters who provide both joy and perspective as he continuously reflects on his parenting journey.
More Speakers
Topic – Machine Learning Life Cycle (MLLC) through a Curious Tester’s Eyes
Loosely speaking, the key steps of a Machine Learning Life Cycle a.k.a. ML Life Cycle addresses Problem definition, Data Requirements, EDA (Exploratory Data Analysis), Feature Engineering and Models (Select, Train, Evaluate, Deployment & Maintenance). There is no rocket science in this life cycle to many and that’s probably true to a good degree.
It is also true that when it comes to actually learning Machine Learning (ML), there is absolutely no dearth of offline and online learning resources for an active, aware and enthusiastic learner. Popular and classical training datasets like Boston House Price, Iris, MNIST and so on are available freely from many credible sources like Kaggle, scikit-learn portal and so on. Numerous code samples have already been published on GitHub in such a way that a beginner and non-expert in ML can easily Clone/Fork It, Run It and Done! One has looked into Machine Learning in Action. But is that all? I don’t think so!
How do we know what’s happening, why is it happening, where and how?
What does it mean what are we doing and what is it producing?
Do we intend to understand reasonably well or we are happy to sign-off learning in a shallow way?
In this proposed demo, I will walk you through the five key steps – that I personally took to get a little deeper into MLLC. I find it useful to continue my journey of approaching learning in slightly more effective and reasonable ways. The five steps are…
- Understand how a typical Machine Learning Life Cycle is generally modelled and described.
- Understand how to zoom-in and zoom out into key phases and steps.
- Understand the Core Concepts, APIs, Libraries, Packages and Techniques used in ML.
- Understand how to work with them like a Thinking-Always-Turned-On Tester throughout a MLLC.
- Reflect, Discuss, and Expand on your learnings using LLMs.
At the end of the demo I hope that you would appreciate how a tester’s curiosity driven learning approach makes a huge difference in how holistically one explores this niche area of learning of what I acronymized as MLLC (Machine Learning Life Cycle).
Target Audience: Software testers, QA / QE professionals, and anyone interested in understanding the basics and fundamentals of Machine Learning and its programming aspects.