Lab: Vector Search Mini Lab

Goal

Create a tiny semantic search system over sample documents.

Tasks

• Split sample-data/rag-documents.md into chunks.
• Generate embeddings or mock embeddings.
• Store vectors in a simple local structure or vector database.
• Retrieve top-k chunks for test questions.
• Compare results with keyword search.

Failure Case

• Ask a question that has no answer in the documents.

Output

• Retrieval script.
• Top-k result table.
• Notes on retrieval quality.