Product
VectorHub is a private knowledge layer for AI applications.
The current repository implements the foundations: document ingestion, collection management, chunking, metadata storage, semantic retrieval and authenticated application workflows.
Available
Retrieval API
`POST /query` is confirmed in the backend and client documentation.
Experimental
Document operations
Upload, parsing, chunking and chunk editor flows exist, but some schemas are still broad RPC surfaces.
Planned
Agent adapters
Agent context is a product direction until a stable SDK or integration contract exists.
Who is this for?
AI developers
Research teams
Product teams
Private infrastructure operators