I'm a Senior AI Engineer and Technical Lead building production agents, MCP infrastructure, evaluation systems, and developer tools.
I work across technical discovery, architecture, backend implementation, observability, and iteration with real users. I care about making AI systems reliable, understandable, and useful beyond the prototype stage.
- Production LLM agents and human-in-the-loop workflows
- MCP servers, clients, gateways, and LLM-optimized tool interfaces
- Evaluation, tracing, and observability infrastructure for AI systems
- Open-source developer tools in TypeScript and Python
| Project | What it does |
|---|---|
| Toolbx | A local MCP gateway that centralizes server configuration across Claude Code, Codex, OpenCode, and other MCP clients. |
| llm-agent-loop | A zero-dependency, provider-agnostic TypeScript runtime for observable and controllable LLM agent loops. |
| ctxman | A terminal UI for managing context files across AI coding agents. |
| agentic-document-extraction | A vision-first FastAPI service for extracting structured data from documents with AI and OCR. |
TypeScript 路 Python 路 MCP 路 FastMCP 路 LangGraph 路 Pydantic AI 路 Vercel AI SDK 路 OpenAI 路 Anthropic 路 Gemini 路 AWS 路 PostgreSQL 路 Docker 路 Kubernetes




