San Francisco Bay AreaI build production LLM agents end to end: the user problem, the multi-agent loop, and the evals that prove it works. Open to AI Engineer and AI Product roles.
Founder habits, engineer's hands. I build agents that ship, and the evals that decide whether they should.
I scope the problem with the people who have it, ship the first version fast, and keep the metric that says whether it helped.
Three published papers: hardware randomness, a streaming-analytics benchmark and distributed GAN training.
Personal projects, built end to end: the problem, the product and the system behind it. Each one opens into its full story.
Best reached by email or LinkedIn. Based in the San Francisco Bay Area.