databricks-metric-views

You can define governed business metrics in Databricks Unity Catalog by creating a YAML specification that separates metric definitions from their usage, enabling consistent KPI reporting. This approach supports complex aggregations and declarative joins for star schemas.

Yes, you can model star and snowflake schemas directly within Databricks metric definitions using declarative joins. This allows you to define relationships for consistent querying across complex data structures without altering the underlying source tables.

Databricks metric views support complex aggregations including ratios, filtered measures, and window measures like YTD or moving averages. You define these calculations declaratively in the YAML specification to ensure consistent execution.

The best way to ensure consistent KPI calculations is to define them as governed metric views within Unity Catalog. This centralizes the logic so Databricks SQL, AI/BI Dashboards, and Genie all query the same standardized definitions.

No external dependencies or tools are required to define business metrics in YAML. The Skill operates independently using references to create and manage metric views natively within the Databricks Unity Catalog environment.

Defining metrics separately in YAML ensures reusable metric definitions that prevent calculation discrepancies across reports. This governance approach centralizes KPI logic so teams always query the exact same source data and aggregations.