I build products from 0 to 1 and systems from 1 to millions.

I’ve spent my career moving between product and infrastructure. I was originally in grad school for data science, but I dropped out and moved to San Francisco to land a job in web development. Since then, I’ve gone deep into databases and distributed systems. I launched a generative AI product that trained custom models on my own GPU. Today, I build applied AI systems supporting more than one million customer interactions each year.

The technology has changed, but the work has stayed largely the same: find the real constraint in an ambiguous problem and build what it takes to solve it in production. A system isn't finished when it launches. It is informed by real usage patterns, changing requirements, and the occasional 2 a.m. outage.

I’m comfortable building from first principles when the standard solution doesn’t fit. That has meant custom ML pipelines, deployment systems, event platforms, and shared infrastructure for AI agents. When I still wrote code line by line, my favorite programming language was Go: simple interfaces and explicit contracts, even if it was a bit verbose. That philosophy shapes how I not just write code but build systems.

I grew up playing team sports, and I still love the same part of software: solving hard problems alongside people with different strengths.

My superpowers are breadth of knowledge and willingness to look dumb while I learn. Both come from a strong intellectual curiosity.