📚 RAG Path¶
L100 L200 L300 L400
Retrieval-Augmented Generation (RAG) is the most common pattern for grounding AI agents in your own data. Instead of relying on the LLM's training data, you retrieve relevant documents at query time and include them in the prompt.
What You'll Build¶
- ✅ Understand the RAG pipeline end-to-end
- ✅ Load, chunk, and embed documents using GitHub Models (free)
- ✅ Store and query vectors with a local pgvector (Docker)
- ✅ Build semantic search over Azure PostgreSQL + pgvector
- ✅ Evaluate RAG quality with the Azure AI Evaluation SDK
Path Labs (7 labs, ~355 min total)¶
| Lab | Title | Level | Cost |
|---|---|---|---|
| Lab 006 | What is RAG? | L50 | ✅ Free |
| Lab 007 | What are Embeddings? | L50 | ✅ Free |
| Lab 022 | RAG Pipeline with GitHub Models + pgvector | L200 | ✅ Free |
| Lab 026 | Agentic RAG Pattern | L200 | ✅ GitHub Free |
| Lab 031 | pgvector Semantic Search on Azure | L300 | Free |
| Lab 039 | Vector Database Comparison | L300 | ✅ Free |
| Lab 042 | Enterprise RAG with Evaluations | L400 | ⚠️ Azure |
The RAG Pipeline¶
