← AI Agents & RAG learning pathAI Agents & RAG · Module 3 of 10
Embeddings and vector search
Search by meaning with embeddings and vector databases.
📖 Key ideas
- Embeddings turn text into vectors that capture meaning.
- Similar meanings give nearby vectors.
- Vector databases store and search embeddings: Pinecone, Weaviate, Qdrant, pgvector, Chroma.
- Similarity measured by cosine similarity or dot product.
- Approximate nearest neighbour (ANN) indexes make search fast.
💡 Real-life example: Embeddings are a map where similar ideas live in the same neighbourhood.
EmbeddingVector representing meaning
Vector databaseStores and searches vectors
Cosine similarityAngle-based similarity
ANNApproximate nearest neighbour search