I'm Sonu. I like finding repetitive problems and turning them into systems.

I started in testing, which is a good place to notice repetition because you are the person doing it. The same regression, the same environment setup, the same test-case document written slightly differently for the twentieth time.

What changed how I work was realising that most of that repetition is not a discipline problem, it is a systems problem. A flaky test is usually an architecture signal. A three-day regression is a statement about how the software is built, not about how hard anyone is trying. Once you see it that way, the interesting work stops being the testing and starts being the system that makes the testing unnecessary.

That took me through mobile and API automation, then into performance engineering, where the question is different again — not does this work, but what happens to it under a load nobody can rehearse. Then into leading a quality organisation, where the same instinct applies to process instead of code.

Now most of what interests me is AI quality: how you test a system that does not have one correct answer, and how much of the repetitive engineering work around it can be handed to the same technology that created the problem.