Agentic AI Buildcamp for Engineers & PMs

If you have never written code or called an LLM API, start with the free lightning lessons. This is also not a course on training models.

Yes, run as a buildcamp: every live session builds the next part of a real system, and every week ends in a project you show at Friday Q&A.

Most teach one shape: a person asks and an agent retrieves an answer. This one covers the map, from fixed workflows to single agents to teams of agents, started by people, events, schedules or other agents, and it measures each one by the outcome it delivers and what that outcome costs.

An inquiry pipeline for a drug maker, where code decides every step and an eval gate guards a safety rule, and AI mission control, where AI employees work from a board, each hired for one outcome with a metric and a rule it must never break.

Both systems, plus your own AI employee on the mission control board: its outcome and metric, the tests that prove it never breaks its rule, its cost per outcome, and a design you defended in a one-to-one architecture review.

Recorded walkthroughs of agentic systems real companies run and have written about, such as expense approval, support resolution, accounts payable and multi-agent research. Each shows the job, the shape on the map, the design, how it is measured and what was traded. Watch them when it suits you.

The sessions use the OpenAI Agents SDK, with MCP for tools. The design carries over to LangGraph, the Claude Agent SDK and Pydantic AI, and we compare them where it matters.

Every session is recorded, so you can catch up before the next one. Bring your questions to Friday office hours, and you can rejoin a future cohort live at no extra cost.

Plan for 4 hours of live sessions each week, the optional 1-hour office hours on Fridays, and 3 to 5 hours of project work.

You can expense it as professional development if your company supports that. Maven applies 20% off automatically for 2 to 9 seats and 25% off for 10 or more.

The list price is $997. Use code FRIENDS300 to get $300 off, which brings it to $697 while enrollment for this first cohort is open. The code closes when the cohort starts.

No. Each session is a live build: I explain the piece, we build it together, and you extend it before the next session.

Agents that plan and act with tools toward an outcome, often several with a coordinator, plus what makes them safe to run at work: permissions, approvals, durable work, cost control, evals and observability. Part of the skill is knowing when a plain workflow does the job better.