INFINITY (by embraceable Technology) icon

INFINITY (by embraceable Technology)

INFINITY (by embraceable Technology) — retrieve from and upload documents to your Knowledge Collections

Actions4

Retrieve

AI-generated

Summary

Retrieve relevant chunks from a selected INFINITY Knowledge Collection using a required query. Choose matching and retrieval settings, then return either one item containing the full API response or separate items for each chunk.

Inputs

  • Knowledge Collection (applicationId) (required) — Select a collection from the list or enter its ID.
  • Query (query) (required) — Search text used to retrieve relevant information.
  • Advanced Options → Include Explanation (includeExplanation) — Defaults to true. Includes the retrieval explanation when available; set false to omit it. In raw output, the node removes the top-level explanation field.
  • Advanced Options → Limit (limit) — Maximum results to return; defaults to 50.
  • Advanced Options → Match Mode (matchMode) — Matching mode: Strict, Balanced (default), or Overview.
  • Advanced Options → Output Format (outputFormat) — Processed (default) fans out each result chunk into its own item and strips the p_ prefix from keys in its fields object. Raw returns the full API response as one item without transformation.
  • Advanced Options → Retrieval Strategy (type) — Retrieval strategy: Agentic Simple (default) or Simple.
  • Advanced Options → Timeout (Seconds) (timeout) — Maximum wait for the retrieval response, in seconds; defaults to 600.

Output shape

A list of items, paired with their corresponding input items.

Raw format returns one item per input with the API response. Processed format returns one item per result chunk, adding retrievalId and applicationId and, when enabled and available, explanation. If there are no chunks, processed format returns one item containing the response, optionally without its explanation. With Continue On Fail enabled, errors are returned as error data items.

Links

Discussion