Build AI agents you can trust in production
Build a real Python agent with structured outputs, tools, tests, and deployment. Learn the engineering around the LLM, not just the API call.
Ten exercises in your browser. No install, no API key.
Self-paced, one-time payment: €99 or €199.
Free. Developers building with AI.
By Bob Belderbos & Juanjo Expósito · 150+ developers coached since 2020
Start free
10 exercises
Structured output, tool use, Protocols, a service layer and an agent loop, tested in your browser.
The challenge
Ten production decisions in 15 minutes: do you build AI systems, or just call an LLM?
The guide
Beyond the LLM Call: one agent, from a 20-line demo to a system you can test. 14 pages.
The course
Six weeks, one real project: Expense AI Agent, built test-first and deployed with Docker. Self-paced, one-time payment.
- Course, €99: the six-week project with 150+ pre-written tests.
- Complete, €199: plus the reference solution, the agent week, advanced exercises and production guides.
Community and live events
An open Slack for people building agents, and occasional free live sessions: one engineering question about AI apps, worked through in code, then your questions.
For companies Custom AI agent engineering → We build agents around your application, integrations and production constraints. Projects from €2,000 excl. VAT; final pricing depends on scope.
Free guide: Beyond the LLM Call
14 pages on the Python system around an AI agent: validation, Protocols, a service layer, human review and a guarded agent loop. Confirm your email to download the PDF, then get the biweekly newsletter on building agentic systems in Python. What's inside →
No spam. Unsubscribe anytime.
