supabase

Why we included this project

Supabase is a good fit when you want semantic search without running a second database. It keeps Postgres as the core and adds the pgvector extension, so embeddings sit in the same tables as your relational data and you can filter them with the SQL you already know. The bundled AI toolkit handles the rest of the pipeline, from generating vectors through OpenAI, Hugging Face, or LangChain to running similarity queries, and a Python client makes managing collections straightforward. Since auth, realtime subscriptions, edge functions, and file storage ship alongside the database, a full-stack app can grow in one place rather than being assembled from half a dozen services. It is also self-hostable, which matters for teams that want the same stack on their own infrastructure.