Vector Store
Represent storage and retrieval support for vectorized content.
Category: Query. Registry label: Vector Store.
Purpose #
A vector store is the conceptual place where embedding representations are saved for later similarity search. In a working RAG system, vector-store configuration must be compatible with embedding dimensions and retrieval needs.
Position in the graph #
Embedder → Vector Store → Retriever or Embedder → Vector Store → Hybrid Retriever
Design guidance #
Make sure the chosen retrieval node follows this stage. Consider the difference between a graph representation of a store and an actual database connection. In the frontend preview, this node is a configurable element rather than a verified live vector database.
Current limitation #
No vector records are persisted, indexed, or queried by the current frontend.