Skip to main content
VectorAI DB integrates with popular AI frameworks, embedding providers, and identity platforms so you can focus on application logic rather than infrastructure. Use any supported integration to generate embeddings, store vectors, and run similarity searches with minimal setup. Choose a framework integration like LangChain or LlamaIndex when you want built-in abstractions for RAG pipelines, retriever chains, and document management. Choose an embedding provider directly when you need full control over how vectors are generated and stored using the VectorAI DB client. Use the Scalekit integration when you need agent memory scoped to authenticated users.

​
Frameworks

LangChain

Use VectorAI DB as a vector store in LangChain for RAG pipelines, similarity search, and retriever-based chains. Supports sync and async operations.

LlamaIndex

Build RAG applications and query engines with VectorAI DB as the storage backend in LlamaIndex.

​
Auth and identity

Scope vector data to authenticated users.

Scalekit

Connect Scalekit per-user identity to VectorAI DB per-user collections so agent memory is scoped to the authenticated user.

​
How integrations work

All integrations follow the same pattern:
  1. Generate embeddings — Use an embedding provider (such as OpenAI or Cohere) to convert your data into vectors.
  2. Store in VectorAI DB — Insert vectors into a collection with optional metadata payloads.
  3. Search — Query with a vector to find semantically similar results, with optional metadata filtering.

​
Quick reference