I build systems that think ahead.

Deep where it counts.

The builds above get the attention. The reason to hire me is what's underneath.

Most of my work is the part users never see. I build backend systems in Python and Node: REST and real-time APIs, the data pipelines that move data between them, and PostgreSQL schemas designed with least-privilege roles instead of a single god-user. Then I run it in production on infrastructure I operate and harden myself: Linux servers, containerized and reverse-proxied, with the firewalling, intrusion detection, and secrets management real systems actually need. No managed platform doing it for me.

AI is where that backend meets the current frontier, and it's where most of my recent work lives. I design agents and Model Context Protocol tooling, wire up retrieval and LLM orchestration, and build computer-vision pipelines: systems that run on a schedule and publish real output, not one-off demos. The kind of thing that has to work on the hundredth run, not just the first.

The visible craft is real too: the browser games, the WebGL, the drawings on this page. But it sits on top of the engineering, not in place of it. I spent years as a designer before I went deep on the backend, which means I can architect the database and make the result look like it cost more than it did. Most engineers only do one of those.