fabric-data-agent

Ground agent answers in Fabric semantic models by routing user queries to governed data sources, fetching verified metrics and dimensions while enforcing workspace permissions and data lineage. This ensures responses come from structured semantic models rather than unstructured sources.

Yes, Power BI datasets work with Fabric lakehouses and warehouses to enforce data governance in agent responses. The system connects to these enterprise sources to return measures, dimensions, and time context with deterministic checks aligned to workspace permissions.

Route metrics questions to governed Fabric data sources by querying semantic models that contain defined measures and dimensions. The agent applies governance rules to return structured data context, ensuring answers are bound to authorized lakehouse and warehouse definitions.

Yes, fabric-data-agent requires existing Fabric workspace definitions to return measures and dimensions. It connects to pre-configured semantic models, lakehouses, warehouses, and Power BI datasets to apply deterministic governance checks and enforce data lineage permissions.

Governed semantic models provide governed analytics by ensuring agent answers come from authorized Fabric data sources rather than unstructured sources. This approach enforces permissions, maintains data lineage, and returns verified metrics with time context for deterministic results.

Grounding agents in Fabric lakehouses and warehouses is limited to data definitions, metrics, and measures available within existing semantic models. It cannot generate answers from unstructured sources or bypass enforced workspace permissions and data lineage governance rules.