AI Campus • Practical resources
RAG & Vector Search Systems
Keep documents traceable while comparing lexical retrieval and local vector search.
Before you start
- Python lists/dictionaries, document chunking basics and a small permitted note collection.
The reading and planning exercises need no paid course or API. Optional inference, training, IDE features and hosted services may require hardware, accounts or paid access. Use the local or manual alternative before spending money.
Start with the official learning material
Free-to-read references checked 3 November 2026. Reading documentation does not activate a provider or create a subscription.
Practice with a concrete exercise
1. Build the retrieval baseline
Follow the cited study-search starter from the related project guide. Write five questions and record the expected source lines. Include one question absent from the collection.
2. Preserve chunks and sources
Read Chroma’s introduction. Define records with chunk ID, document ID, text and page/line metadata. Plan what must be rebuilt when the embedding model or chunking strategy changes.
3. Compare top-k retrieval
Optionally create a local Chroma collection using a compatible local embedding model. Compare top-three results against your baseline and manually labeled support passages. Evaluate generated answers only as a separate extension.
Check what you learned
- Returned chunk IDs resolve to the exact original text.
- Unsupported questions are included in the test set.
- A search result is not labeled a verified generated answer.
Keep your artifact, observed checks and remaining limitations in a short README. Mark any exercise you only designed rather than executed.