About Me

I build AI agents that work in the real world, under real constraints.

I’m a Senior Lead Machine Learning Engineer at Capital One, where I work on agentic AI: retrieval and search systems, vision-language models, and the evaluation and observability layers that make LLM systems safe to ship in a regulated environment.

Over the past decade I’ve built production ML across fintech, media, health tech, and live events, at Capital One, Headspace, Ticketmaster, and others: recommendation systems, A/B testing platforms, LLM pipelines, and ML infrastructure. Along the way I’ve co-authored a patent, published research, and won the RBC & Microsoft Hackathon.

What I write about here

This site is where I work through how large AI systems actually function, usually by building them or animating them. So far that’s meant training a language model from scratch and how training is parallelized across GPUs. Next up: agents, evaluation, and LLM inference.

Get in touch

I’m always happy to talk about agents, evals, and LLM inference, or to help engineers moving into AI.

• LinkedIn
• GitHub

Views expressed on this site are my own and do not represent my employer.