Vector Database
A store built to index and search embeddings by nearest-neighbor similarity at scale.
A vector database indexes high-dimensional embeddings and answers “which vectors are closest to this one?” fast, using approximate nearest-neighbor indexes (HNSW, IVF). It’s the retrieval backbone of RAG and semantic search.
Worked example: a database that stores embeddings and answers ‘find the k nearest vectors to this query’ using an approximate-nearest-neighbor index (HNSW, IVF), trading a little recall for a huge speedup over brute force. Gotcha: it is an ANN index, not a relational store — for small corpora (under ~100k vectors) a plain array + cosine loop or pgvector is simpler and exact; reach for a dedicated vector DB when scale, metadata filtering, and hybrid search justify it, not by default.