Traditional search matches exact words, which is useful when the exact term matters. Vector search finds results by meaning and semantic similarity.
A vector database stores and searches embeddings efficiently. Examples include pgvector, Pinecone, Qdrant and Weaviate.