I build production infrastructure for AI agents.
My work focuses on three areas:
- Agentic RCA — evidence-driven incident diagnosis across metrics, logs, traces, and source code
- AI Gateway — unified model routing, protocol compatibility, streaming, usage, and reliability
- MCP Infrastructure — tool discovery, authentication, authorization, credential isolation, rate limiting, and audit
I work primarily with Go, Python, Kubernetes, LangGraph, MCP, and OpenTelemetry.