Skip to main content

Semantic Query

Overview

The Semantic Query control enables natural-language access to structured data by translating user intent into executable queries. Powered by Large Language Models (LLMs), it automatically generates and executes SQL or API parameters against connected data sources—allowing users to retrieve precise results without understanding schemas or query syntax.

Configuration

To configure a Semantic Query control, define both the Brain (reasoning model) and the Body (data source).

1. Model Configuration

PropertyDescription
NameUnique identifier for this Semantic Query instance
DescriptionContextual guidance for the model about the query’s purpose
ModelLLM provider and version used for reasoning
TemperatureControls determinism vs. creativity in query generation
Connector LibraryData connector used to execute generated queries

2. Data Source Configuration

  1. Select a connector from the Connector Library.
  2. Choose the target table or API resource.
  3. Provide schema metadata to improve semantic accuracy.

Once configured, the Semantic Query control is ready to accept natural-language prompts and return structured results.