Application developers who want RAG and semantic search built directly into their PostgreSQL database.

A Timescale toolkit that adds automatic embedding generation, a background vectorizer, and LLM function calls directly inside PostgreSQL via SQL.

Installed as a PostgreSQL extension and worker process alongside an existing Postgres database.

Use when a team wants RAG or semantic search without standing up a separate vector database or embedding pipeline.

It keeps embeddings automatically in sync with source tables so retrieval logic can live entirely in SQL.

Works alongside pgvector for storage and calls out to external embedding and LLM providers.

Open source under the PostgreSQL license, maintained by Timescale; free to self host or use via Timescale Cloud.