Build a Fabric data agent with an ontology
Beginner
Data Analyst
Data Engineer
Microsoft Fabric
Create a Fabric data agent that uses a Microsoft Fabric IQ ontology as its data source. Configure agent instructions to improve query accuracy, test natural language questions, and publish the agent so business users can get governed data answers without writing code.
Learning objectives
By the end of this module, you'll be able to:
- Explain how an ontology provides business vocabulary that a Fabric data agent uses as a data source
- Create a Fabric data agent and connect it to an ontology as a data source
- Configure agent instructions to improve query accuracy
- Test data agent responses by asking natural language questions and reviewing generated queries
- Publish a Fabric data agent to make it available to colleagues
Prerequisites
Before starting this module, you should have:
- Understanding of ontology concepts, including entity types, properties, relationships, and data bindings. If you're new to ontologies, complete the Create an ontology with Microsoft Fabric IQ module first.
- Familiarity with Microsoft Fabric workspaces and navigation