pgvector is a Postgres extension that adds vector storage and similarity search to an existing database, so you can run semantic queries directly against your application data without standing up a separate vector store. If you're already on Postgres, you can enable it with one CREATE EXTENSION statement, add a vector column to any table, and have semantic search returning results the same day.

This post walks through adding it to an existing app — from installing the extension to running your first semantic query, with an HNSW index for performance at scale.

TL;DR

What it is: A Postgres extension that adds a vector column type and similarity-search operators ( <=> , <-> , <#> ).

Why it matters: Semantic search without a separate vector database, hybrid keyword-and-semantic queries in one SQL statement, and no new service to operate.

Five steps to ship it: Install the extension, add a vector(N) column, embed at write time, query with cosine similarity, add an HNSW index for scale.

Embedding cost:…