I build multi-agent systems and the data infrastructure underneath them — fraud investigation agents, streaming lakehouses, and the occasional Kubernetes cluster upgrade at 2am. Off the clock: Indian forts, dynasty timelines, and weekend cricket.
I work at UIDAI, where my path has moved from building the data foundation to leading work around agentic AI. I have worked on a roughly 400 TB data platform, real-time streaming, lakehouse infrastructure, risk scoring and a population-scale feature store — systems where correctness, scale and auditability matter as much as raw engineering speed.
More recently, I have been building Prahari, a multi-agent investigation system designed to bring evidence, case history, internal tools and runbooks into one workflow. The goal is not "AI for the sake of AI"; it is to reduce repetitive analyst work while keeping decisions grounded, traceable and reviewable.
Outside work, I am increasingly interested in the space between technology and institutions: digital governance, public-sector consulting and eventually an MBA or public-policy path. I also travel, read history, collect stories about Indian forts and dynasties, and keep a few ideas alive on the side.
I like visiting old places and then going down the rabbit hole: who built them, why they mattered, how borders shifted, and what survives in the architecture. Recent trips and plans have included Orchha, Maharashtra's temple-and-fort circuit, South India and Europe.