Polygres

Polygres is an all-in-one PostgreSQL platform for AI agents, combining relational data, graph traversal, vector search, and hybrid retrieval behind a single unified API. Instead of juggling a separate vector database, a graph database, and a fragile synchronization pipeline, you can turn the PostgreSQL databases you already run into working memory for your agents.

With Polygres, agents can retrieve structured rows, follow connected relationships, surface semantic matches, fetch full-text results, and pull in filtered context — all from one place. That means less infrastructure to maintain and fewer moving parts to break when your agents need the right answer at the right moment.

The platform ships with two purpose-built extensions: pgGraph handles graph traversal based on foreign keys, so agents can navigate relationships stored naturally in your tables, while pgContext delivers fused retrieval across dense HNSW vectors, sparse search, full-text search, filtering, recommendations, discovery, grouped results, and lookups.

You can deploy Polygres as a managed cloud database when you want things to just work, or self-host its open-source components when you need full control over your data and infrastructure.