A 6-week, hands-on program on building and deploying secure AI agents.

Who this is for

Working professionals — engineers, PMs, founders, consultants — who want to move past chatbots and run a real, always-on agent against their own work or job search. You should be comfortable in a terminal, willing to read API docs, and ready to break things in public.

You don't need a research background. You do need to come prepared to build, ask specific questions, and share what you ship.

What you'll do

By Week 6, every participant who wants to will deploy their own personal AI agent — and will know what they've trusted it with, how to scope its permissions, and when to pull the plug.

You'll either run a managed agent (BareClaw, a one-click cloud OpenClaw deployment) or self-host with OpenClaw / NanoClaw on your own hardware. Both paths are fully supported.

How sessions work

Mondays, 10:30 AM PT / 1:30 PM ET, 90 minutes. Slides + live demo + Q&A + peer breakout. Weeks 1–2 are foundations, Weeks 3–4 are build labs, Weeks 5–6 are lightning talks and the member showcase. All session materials — slides, recording, references — land on the Sessions page after each week.

Real-time questions between sessions happen in the cohort Slack. Self-serve invite — admin approves each join. Per-session async comments live under each session page.

A note on the venture

BareClaw is built and operated by members of the program team. It's a real product — it costs real money to run secure containerized agents in the cloud — and it will come up a lot in Season 2 as both a lab and a live case study.

We're calling that out upfront because you signed up for a peer learning experience, not a sales funnel. If the managed path isn't for you, the program is still free, the peer learning is real, and you can build toward the same north star on your own.