I build AI infrastructure that is meant to hold up beyond the demo.
My current work centers on autonomous coding agents, operational memory, and the tooling needed to keep AI-assisted software development reviewable, repeatable, and useful in real projects.
- SynapseGrid Labs - my AI systems practice for autonomous agents and automation infrastructure
- BrainCore - operational memory infrastructure for AI systems using PostgreSQL, pgvector, and TypeScript
- Agent workflows - GitHub-native patterns for autonomous build agents, private workspaces, and public release hygiene
TypeScript, Python, PostgreSQL, pgvector, Docker, GitHub Actions, Cloudflare, Linux, browser automation, and local/hosted AI tooling.
I care about systems that can be inspected and operated: clear interfaces, durable records, practical automation, and enough verification that agents can work without turning the codebase into a black box.