AI Engineer Roadmap
Build agents that run unattended, then prove it in a case interview. The most complete path on the site.
Billed every year. One subscription, 35 courses, every program included.
This AI engineer roadmap puts the courses in the order you'd learn the job. First you set up your own working environment and learn Claude Code. Then you build: retrieval, tool calling, agents and evals. Last, you make the work visible and prepare for interviews.
It suits developers moving into AI engineering. If you don't code yet, the foundation stage includes Claude Code for Absolute Beginners, so you can start from zero, but expect the build stage to take you longer.
Foundation
5 courses · 61 lessonsCore skill building
2 courses · 134 lessonsHands-on build
3 courses · 148 lessonsGet hired
4 courses · 41 lessonsWho the AI Engineer Program Is For
- Developers who want to move into AI engineering and build with LLMs
- Engineers who want to build AI agents that run without supervision
- Anyone preparing for AI engineer or agent engineer interviews
What You Will Be Able to Do
- Build a production-shaped LLM application with retrieval (RAG), tool calling and MCP
- Design, build and evaluate AI agents on realistic cases
- Measure AI quality with evals instead of guessing
- Explain your design choices in an AI agent case interview
AI Engineer Program FAQ
Do I need coding experience for the AI Engineer program?
The core courses are written for junior developers. If you have never coded, start with Claude Code for Absolute Beginners, which is part of this path.
What should I learn first to become an AI engineer?
Working with models through their APIs, then retrieval (RAG), then tool calling and evals. Agents come after those, because an agent is those pieces put together. The roadmap above follows that order.
Does it cover AI agents?
Yes. Agents run through the whole path: an agent built from first principles, fifteen hands-on agent cases, and an agent design interview course.
What do I have to show at the end?
Built projects you can put on GitHub and talk through in an interview, plus a rehearsed live demo of your work.