ChatGPT plugins now support Postgres & Supabase

Summary

Retrieval plugins let ChatGPT access relevant content from private datasets by chunking documents, generating embeddings with OpenAI’s text-embedding-ada-002, and querying a vector database. The article describes Postgres and Supabase datastore implementations using pgvector, with a shared documents schema and a match_page_sections function that filters by source, author, document, and date before ranking vectors with inner product distance. It then gives a local example: downloading the Postgres 15 documentation as a PDF, uploading it through the plugin’s /upsert-file endpoint, and configuring environment variables for Supabase or Postgres. The example produced 7,904 document records, and an optional IVFFlat index can trade exact nearest-neighbor search for approximate search; after registering the local plugin in ChatGPT, users can ask questions answered from the documentation.

Takeaways

1. 1

   The Postgres datastore uses psycopg2 for direct database access, while the Supabase datastore uses PostgREST, which supports Row Level Security and use of the data beyond ChatGPT.

2. 2

   The shared match_page_sections function filters records by source, author, document, and date, then finds nearest embeddings with the inner product distance function and returns a similarity score.

3. 3

   Uploading the Postgres documentation created 7,904 document records. An IVFFlat index can speed searches, but changes exact nearest-neighbor search to approximate nearest-neighbor search.