Lesson 2 of 8 · Retrieval Basics: Make AI Use Your Own Documents
In this lesson. Split on structure, overlap a little, keep citations.
What you will learn
- Structure-aware splits
- Overlap
- Source ids
Walkthrough
You will chunk by heading, keep overlap, and store source ids. Bad chunking is why retrieval “does not work.” This lesson is mechanical and important.
Work through the ideas in order. After each point, pause and connect it to a task you already do — a document, a workflow, or a feature you own. The goal of Retrieval Basics: Make AI Use Your Own Documents is usable skill, not a pile of notes.
If something is unclear, rewrite it in your own words before you continue. Teaching the step back to yourself is the fastest way to see gaps.
Practice
Chunk one article into 8–12 pieces with source ids you could show a user.
Keep the first attempt small. A finished example you can reuse beats a perfect plan you never run.
Check your understanding
- Can you explain the goal of this lesson in one sentence to a teammate?
- Where would you apply “Structure-aware splits” in your own work this week?
- What would you change on a second pass of the practice?
Next. Continue to the following lesson when the practice has a real artifact, even a rough one.