Lead Data Scientist at Urban Company with ~7 years in applied ML, currently leading production GenAI and multi-agent systems at consumer scale. I lead applied LLM and multi-agent work end-to-end — designing agentic, real-time applications that ship and hold up under real-world load. I'm currently building and have deployed production grade voice agents with multi-agent orchestration, large-scale computer-vision and OCR pipelines processing millions of images a week, and the classical ML and decision-science systems underneath them — forecasting, causal modelling, fraud detection, and supply-chain optimisation. Across all of it I care about the same thing: systems that move a real metric and survive contact with production. My foundation is in operations research and supply chain — a B.Tech in Production & Industrial Engineering from IIT Roorkee — which still shapes how I reason about optimisation and constraints in the systems I build. Always interested in hard problems at the intersection of LLMs, applied ML, and systems that operate at scale.