Recherche Semantique

Vector similarity search in the knowledge base. Returns the documents most semantically similar to the query.

Parameters

Parameter	Type	Required	Variable	Description
query	dynamic value	Yes	Yes	Natural language question to search the ontology.
synthesize	boolean	No	No	Enable AI model synthesis of results. (Default: true)
explain	boolean	No	No	Include a reasoning explanation in the response. (Default: false)
model.provider	text	No	No	AI model provider for synthesis (e.g. openai, anthropic).
model.model	text	No	No	AI model identifier to use for synthesis.
temperature	number	No	No	Model creativity for synthesis (0 = deterministic, 1 = creative). (Default: 0.7, min 0, max 1)
maxTokens	number	No	No	Maximum number of tokens in the synthesized response. (Default: 2048, min 1, max 128000)
responseFormat	choice (text, json, structured)	No	No	Response format: plain text, JSON, or typed structure. (Default: "text")
outputVariable	text	No	No	Output variable name containing the search result.

Parameters marked Variable = Yes accept the {{blockName.field}} syntax.

Output

Output variable : searchResults

{
  "results": [],
  "count": 0
}

Example

Search for documents similar to a question.

Input :

{"query": "politique de remboursement"}

Output :

{"results": [{"id": "doc-1", "content": "Notre politique de remboursement...", "score": 0.92}], "count": 1}

Tip

{{searchResults.results}} contains an array of objects with id, content, and score. Adjust topK to control the number of results (default 5).