Hi Bobby,

Tomorrow, I’m building a production AI agent live.

Not a toy chatbot.

Not another "here's how tool calling works" tutorial.

We're going to build the whole thing together: a router agent that decides or denies before the expensive model ever runs, structured outputs with Zod, guardrails against prompt injection, token cost on screen, tracing, and evals with an LLM as the judge.

This is one of the most high value skills right now and I'm shocked how few developers have been exposed to it.

Show up, build with me, break some stuff, and by the end you’ll (hopefully) have an actual working agent you can keep messing with afterward.

Can’t make it live? Register anyway and I’ll send you the replay.

See you tomorrow,

- Brian