About

I'm an AI/ML engineer in New Delhi, freshly graduated with a B.Tech in Industrial IoT — a degree that spent four years teaching me what happens when clean theory meets sensors that lie.

My skill set sits in three places. Deep learning on signals — PyTorch, CNNs and autoencoders, the kind of models that learn what normal looks like so they can flag what isn't. Generative AI infrastructure — LangChain, RAG, vector search, and the evaluation and cost tooling that decides whether an LLM feature survives a budget review. And the production half — FastAPI, Docker, CI/CD and drift monitoring.

I gravitate to the unglamorous side of ML: the caching, the test harnesses, the alerting. It's usually what separates a good model from a good product. I'd rather own a system end to end than a slice of one, and I'll build the tool myself when the existing one hides what it's doing.

what's next A team that puts models in front of real users — where I can take something from first experiment to deployed service, and stay responsible for it after it ships.