Databricks adds Power BI publishing for shared business metrics in private preview
The integration translates metric definitions into Power BI reporting models. Live queries retain Unity Catalog controls, while automated publishing carries definition changes into reports.
Databricks announced Publish to Power BI in private preview on November 10, 2026, letting teams turn Unity Catalog metric views into Power BI semantic models while keeping business definitions in Databricks. DirectQuery is the default and runs calculations in Databricks, where Unity Catalog controls apply; Import mode is also supported, but the stated query-by-query enforcement applies to DirectQuery. Teams can publish manually or automate updates through Lakeflow Jobs, while metric-view materialization is described as a coming optimization, not a current preview feature.
01
Published metric views map dimensions to Power BI columns and translate measure expressions into DAX measures.
02
The preview supports aggregations, ratios, filtered and composed measures, time-window calculations, and field formats such as currency and percentage.
03
Lakeflow Jobs can publish changes on a schedule or after upstream tables refresh, keeping Power BI models aligned with updated definitions.
Power BI users can build reports from business metrics defined in Databricks, rather than recreate those calculations inside their reporting tool. Databricks introduced Publish to Power BI on November 10, 2026, in private preview. The integration turns Unity Catalog metric views into Power BI reporting models while keeping the original definitions in Databricks.
The intended payoff is consistency across reports and AI tools. Databricks says the same metric views can supply Power BI, its AI/BI dashboards and Genie. A business definition—what counts as revenue, for example—can live in one place instead of being maintained separately for each tool.
Business definitions become report fields
Metric views package business calculations with dimensions: the fields, such as region or campaign channel, used to break down a result. Publishing translates that structure into a Power BI semantic model, the model behind its reports. Measures become DAX measures, using Power BI’s calculation language.
Power BI queries the metric view rather than rebuilding its relationships with underlying tables. Joins and calculations run in Databricks. That means a dimension drawn from a joined table appears as an ordinary report column, without requiring the report author to recreate that relationship in Power BI.
The preview covers more than simple totals. Databricks lists support for several calculation types and carries descriptions and display formats into Power BI field properties:
Standard aggregations, including sums, averages, medians and distinct counts.
Ratios, filtered measures and calculations composed from other measures.
Current, offset and trailing-window measures, alongside currency, decimal and percentage formatting.
Live-query governance is mode-specific
DirectQuery is the default mode, sending live queries to Databricks SQL or Lakehouse//RT. Databricks says those queries honor Unity Catalog access controls and security policies, including row filters and column masks. The integration also supports Import mode, which loads data into Power BI; the announcement’s promise of query-by-query Unity Catalog enforcement applies to DirectQuery.
One performance feature remains ahead. Databricks says Power BI will soon benefit from metric-view materialization: precomputed results that can answer a matching query without recalculating it. When no stored result matches, the query would run against the metric view as usual. The company says this optimization would preserve the metric definitions and results.
Definition changes can travel through a pipeline
Teams can publish on demand from Catalog Explorer or add a Power BI task to Lakeflow Jobs, Databricks’ pipeline orchestration system. That task can run on a schedule or after upstream tables refresh. Updated measures, new measures and new dimensions then flow into the target Power BI model through automated publishing.
Databricks illustrates the workflow with marketing reports. A team defines spend, revenue and return on ad spend by channel and region. If it later changes what counts as revenue, it updates the metric view and republishes through its pipeline, rather than editing that definition separately in each Power BI report.
Private preview access runs through Databricks account teams. Once enabled, users prepare named fields and measure inputs, sign in with Power BI credentials, and select a target workspace. Databricks’ final setup step is validation in Power BI: check that the published table, columns and measures arrived as expected.
Sources
databricks.comUnity Catalog Metric Views, now in Power BI: one unified, governed semantic layer for all your reporting and AI
Reader comments
Newest comments first. Replies stay oldest first.