Free textbook from a Google engineer: 400 pages to build an AI agent that works for you
September 2025 — Senior Google engineer Antonio Gulli stirred the developer community by announcing the release of a free textbook: “Agentic Design Patterns: A Hands-On Guide to Building Intelligent Systems.” It’s nearly 400 pages of structured knowledge, covering everything needed to build an autonomous AI agent — from setting up basic processes to fine-tuning behavior via advanced prompting. The author demonstrates how to train an agent to use a browser, connect external APIs, retain memory, set goals, and plan actions.
The textbook spans more than twenty agent design patterns, enabling a shift from simple chatbot experiments to serious systems capable of replacing humans in routine digital tasks. The intrigue: it’s not only aimed at engineers but also at enthusiasts who want to learn how to build AI systems from scratch.
Gulli promises a release in December 2025, emphasizing free access. The exact distribution format remains unclear — whether it will be fully open source (with rights for commercial use of code examples) or under a limited license. Still, the very existence of such a comprehensive, public work is impressive.
At a time when Europe and the U.S. are debating regulations and risks, and corporations are experimenting with embedding AI into browsers and office suites, here comes a textbook that could turn any motivated user into the builder of their own neural agents.
This is not just another guide — it reflects a cultural shift in AI: from consuming ready-made services to creating one’s own tools, where anyone can become the author of a digital assistant. Preliminary materials are already available via Google Cloud’s official resources and the author’s announcements.
You can learn more here.
