Module 1: RAG Foundations
1 min read·13 Oct 2026
What you need first
- Basic Python: lists, dictionaries and functions
- A Google account for Colab, and a free Groq or Gemini key (course-guide/02)
- No earlier module: the course and the
hollin-answersrepository start here
What you will build
The hollin-answers repository with its hollin/ package, the first grounded answer to Ellie with checked citations, and a written decision on which of Hollin's sources are retrieved and which are looked up live in the case database.
Lessons
| File | Title | Minutes | Depth tags |
|---|---|---|---|
01-why-rag-exists.md | Lesson 1: Why a model needs your documents to answer | 40 | Essential |
02-anatomy-and-failure-modes.md | Lesson 2: Nine stages, and where answers quietly go wrong | 50 | Essential, Recommended |
03-choosing-rag-or-something-else.md | Lesson 3: Choosing RAG, long context, fine-tuning or tools | 45 | Essential, Recommended |
04-the-course-stack.md | Lesson 4: The course stack: every tool, one question | 45 | Essential, Recommended |
Extras
05-module-lab.md: Lab: Start Hollin Answers with one cited answer (60 minutes): the package, a cited answer with its check, the per-source decision, a case lookup, six keyless tests and the first commit.06-incident-file.md: Incident: Search answers that cited the wrong source (30 minutes): the Tow Center's 2025 test of eight AI search tools, replayed at Hollin and fixed with a citation gate.