Module 1: Prerequisites
1 min read·13 Oct 2026
What you need first
- Python: functions, classes, exceptions, and reading a stack trace.
- Calling a language model from code, and asking for structured output.
- What embeddings and basic retrieval (RAG) are for.
- The course setup done, with a Groq or Gemini key, or Ollama on your own computer.
If any of these is new, start with the AI Engineering course, Modules 7 and 8.
What you will build
The start of the returns_desk/ package: the shared data, one validated model call that classifies a return request with retries and a cost log, a help-centre search, and tests that run with no key.
Lessons
| File | Title | Minutes | Depth |
|---|---|---|---|
| 01-engineering.md | Lesson 1: The plain engineering underneath every agent | 40 | Essential |
| 02-llm-basics.md | Lesson 2: Talking to a model: calls, tokens, schemas, retrieval | 45 | Essential |
Extras
- 03-module-lab.md: build the returns desk's first working slice, from repository layout to tests, about 60 minutes.
- 04-incident-file.md: how a units mismatch at one interface lost Mars Climate Orbiter, replayed on BrightCart's refunds, about 30 minutes.