Where to start, and why
The same four courses, ordered three ways depending on what you are trying to do. Nothing is locked — take any course in any order. These are suggestions with reasoning attached, not prerequisites.
Applied LLM Engineering
2 courses · 26 minGet useful, reliable output out of a language model, then put it into production.
Prompting first, because everything downstream is a prompt with more machinery around it. Agents second, once you can already make a single call behave.
See the routeAgentic Systems
3 courses · 34 minBuild systems that plan, call tools, and recover from their own mistakes.
The longest route, and the one that needs the most foundations: prompting, then orchestration, then the architecture underneath so failures are diagnosable rather than mysterious.
See the routeAI Safety & Governance
2 courses · 26 minShip models responsibly, and be able to say why your system behaves as it does.
Deliberately short. You need enough hands-on understanding to know what you are governing, and then the governance itself — not a full engineering path first.
See the route