Retrieve Edge

Vector-based retrieval is a core component of the RAG (Retrieval-Augmented Generation) workflow. It enables your system to retrieve the most semantically relevant content from previously indexed knowledge blocks by comparing vector embeddings. You can choose different embedding models in settings.

Retrieve Edge in Workflow

Once your documents are indexed into a collection, you can use the Retrieve edge to find relevant content.

Inputs:

A query block: The text or structured query from user or workflow.

A document block: The indexed structured content (chunks with embedding vectors).

Outputs:

The top-matching structured chunks from the database.

Indexed Structured Data (Vector DB):

Choose the indexed dataset for search.

Configuration:

Result number (Top-K): Number of top results to return (default: 5).

Similarity Threshold: Minimum similarity score (0–1). Default is 0.7. Results below this value will be filtered out.