← All projects
intermediate5h
Semantic Search Mini-App
Chunk a small set of documents, embed them, and build a similarity search over the results — the foundation of retrieval-augmented generation.
Tracks: AI, LLMs, RAG & Agents
Objectives
- Split documents into retrieval-sized chunks
- Represent chunks and queries as vectors
- Rank chunks by cosine similarity to a query
- Explain the limits of pure semantic search versus hybrid search
Before you start
Milestones
0%