I’m a serial founder who builds AI agents for high-stakes work. Today I work on agents at Sully, which gives health systems more capacity for care, from the front desk to the visit and everything after. I also build Kernel, which helps the Department of War find and field technology that already came out of university research.

The question running through all of it: as agents take on longer, higher-stakes work, can the people responsible still trust, inspect, and steer what they did? That is also the idea behind Merton, an AI research manager deployed in Stanford labs.

With Andreas Haupt I co-authored Bruno: An AI Product Manager for Scientists, awarded in the AI Scientist Proposal Competition at the ICML 2026 AI for Science Workshop, and Prototyping Intelligent Organizations, about Giuseppe, an AI foreman that ran a three-person team for three weeks.

Earlier I was the founding product engineer at Compute Exchange, designing the auction and RFQ systems behind price discovery for $1B+ in GPU supply. I spent two years in Rio de Janeiro at Loon Factory, building travel features used by 20 million Brazilians and citizen-services software for the state government.

I studied product design at Stanford’s d.school with an AI focus, advised by David Kelley. Before that I won first place in Environmental Engineering at Intel ISEF for work on pesticide biodegradation, and MIT Lincoln Laboratory named a minor planet after me. I speak English, Hindi, Tamil, and Portuguese.