👋 Hi, I'm Ethan Miao

I'm a senior software engineer and tech lead with 10 years of experience building IoT platforms — device management, cellular connectivity, billing, identity — and for the past year and a half, LLM agent systems running in production.

This blog is where I write down what production teaches me: the architecture that survived, the bugs that almost didn't let it, and the lessons that only show up after the 100th turn of a session in week 40.

Things I write about:

• AI agents in production — LangGraph, multi-agent orchestration, RAG, evaluation, observability
• Backend systems — Java/Spring, Python/FastAPI, Go, and the messaging/storage glue that holds platforms together
• The unglamorous parts — checkpoints that overflow, tokens that expire, streams that leak, and how to design for them

Start with the latest: 5 Things I Learned Running LangGraph Multi-Agent in Production for a Year, or browse the full archive. More about me here.