I build AI-native developer tools, research systems, and practical infrastructure.
Most of my work sits between AI research and systems engineering: agent workflows, model gateways, evaluation tooling, browser automation, MCP tools, monitoring agents, and local-first utilities.
TypeScript / React / Node.js · Go · Python · Rust
PyTorch · Docker · PostgreSQL · Redis · GitHub Actions
- AI agents, tool use, and workflow automation
- LLM gateways, routing, provider operations, and observability
- Reproducible research tooling for evaluation, privacy, and safety experiments
- Systems that are easy to run, inspect, and maintain