AI RAG and pgvector DB Info

[ILoveHostasaurus]

Any version of Empresa whose distribution download included a lib/databases/postgresql.so file is compatible; I believe it's been supported for many years. A database lib method of postgresql must be defined in its config and pointed at that postgresql.so file, just like MySQL. A system lib code of pq must be defined in its config and pointed at the local operating system libpq.so.

[MLT]

Thanks! We do have a need. Chose this approach over Qdrant. Is there a Empressa version required for PostgreSQL? Can you provide any additional information?

-mike

[ILoveHostasaurus]

Our use of pgvector is pure SQL/postgres, there have not been new MivaScript functions introduced. Empresa does support PostgreSQL though, if you did have a need to interact with with a pgvector-enabled postgresql schema.

[MLT]

Hi,

I noticed in a bug-fix listing for Miva 10.13.00 there was refence to vector database support, specifically pgvector. We are integrating RAG into one of our AI projects and would love any information of mivascript's built in pgvector functions. Are there any other AI related functions that are available?

We've ported a good part of our AI to mivascript as the orchestrator of multiple AI agents in a primarily deterministic model, ergo heavy reliance/involvement of RAG.

The Vexture product/feature looks leading the pack! Not surprised Miva team!

Thanks!
-mike