databricks-metric-views

You create governed metric views in Databricks by defining reusable measures and dimensions in YAML syntax. This separates metric logic from data structure, enabling standardized KPIs across dashboards and AI tools via Unity Catalog.

Databricks metric views are governed SQL objects defined in YAML that standardize business metrics. They solve inconsistent KPI reporting by storing reusable definitions in Unity Catalog, eliminating duplicated SQL logic across teams and tools.

You query governed metrics in Databricks SQL using the MEASURE() function. This allows flexible aggregation of defined measures, including calculating complex ratios and window measures without rewriting underlying SQL code.

Yes, the Databricks metric views YAML syntax requires Databricks Runtime 17.2 or higher. You also need active SQL warehouse permissions to properly create, manage, and query the metric views stored in Unity Catalog.

Yes, you can use a single revenue metric view across multiple BI dashboards and AI tools. Storing metric views in Unity Catalog enables shared cross-team access, ensuring consistent naming and logic without rewriting SQL.

You should use Unity Catalog metric views instead of raw SQL to enforce cross-team governance. This approach standardizes metric definitions in YAML, preventing duplicated logic and inaccurate analysis while allowing flexible querying.