Unity Catalog Metric Views, now in Power BI: one unified, governed semantic layer for all your reporting and AI
Databricks introduces Unity Catalog metric views integration with Power BI, enabling a single governed semantic layer for consistent business metrics across reporting and AI tools.
Organizations often face inconsistencies in business metrics like revenue or customer counts across reports and AI applications due to fragmented definitions. Unity Catalog metric views address this by allowing customers to define key performance indicators once in Databricks within an open, governed semantic layer atop the lakehouse. This ensures all teams and AI systems reference the same metrics, reducing discrepancies and improving decision-making. The new integration with Power BI extends this consistency to business users who rely on the tool for reporting.
Many enterprises use Power BI as their primary front end for business reporting, but managing thousands of workspaces with separate semantic models creates governance challenges. Each team’s custom metrics and relationships lead to duplicated logic, causing reports to diverge over time. The new Publish to Power BI feature enables teams to connect Unity Catalog metric views directly to Power BI semantic models, ensuring all reports reflect the same governed definitions without manual rework.
With Publish to Power BI, each metric view becomes a table in the semantic model, where dimensions translate to columns and measures to DAX expressions. Power BI queries the metric view itself, preserving the source of truth and eliminating the need to rebuild relationships in Power BI. For example, a marketing team can define a campaign performance metric view with dimensions like channel and region, then publish it to Power BI for self-service reporting while maintaining alignment with AI/BI dashboards and Genie.
The integration also supports automation through Lakeflow Jobs, allowing teams to schedule or trigger updates to Power BI semantic models whenever metric views change. Future enhancements include materialized views for performance optimization and real-time querying via Lakehouse//RT + DirectQuery, ensuring Power BI reports remain fast and governed. This approach centralizes metric definitions in Databricks while preserving Power BI’s usability for business users.