It started at IIT Bombay. I walked in expecting to study, and walked out having caught a bug that never left me: the pull of machines that think. The moment I saw a circuit obey code — sensors reading the world, motors reacting — I was hooked.

That fascination became robotics. Self-balancing robots, drone navigation, terrain-crawling bots — I chased the messy, physical problems where software meets the real world and nothing behaves the way the textbook promised. Every failed prototype taught me more than a working one.

But robotics kept dragging me back to one thing underneath it all: mathematics. Control systems, linear algebra, probability — the elegant machinery that made the machines actually work. The more I built, the more I fell for the math behind the behavior.

That love of math + real-world messiness had an obvious next home: data science. Instead of teaching a robot to move, I taught models to decide — predicting disease risk, detecting faults, forecasting the unpredictable. Same instinct, new medium.

And then came Generative AI — the frontier where the models don't just predict, they create. NLP, LLMs, systems that reason and generate. It's the same kid staring at a circuit, just with a much bigger playground. I'm still chasing the same feeling — and still one pip install away from the next rabbit hole.