CourseAI Engineering: The Complete Course · Module 8: Retrieval-Augmented Generation (RAG) · part 60 of 105

Part 60 · Module 8: Retrieval-Augmented Generation (RAG)

Lesson 1: Getting the help centre in, cleanly

20 min read·13 Oct 2026

Module 8 · Lesson 1 · About 40 minutes · Needs: Module 4 (what an embedding is), Module 5 (keyword search over the help centre)

The situation

In Module 5 you searched BrightCart's twelve help-centre articles as a tidy Python list. Then Priya sends the real export: the returns pages are Markdown, delivery is a web page, payments a Word file and the warranty guide a PDF.

Before the assistant can answer from them, someone has to get the words out, cut them into pieces a search can return, and label each piece so an answer can cite it. That work is ingestion. Every RAG system (retrieval-augmented generation: search first, then answer from what was found) rests on it.

The rest of this course is yours to keep

This course is bought on its own, once, and stays readable afterwards, including the parts added to it later.

Buy this course · ₹399 See what is inside