Feature Views
GALast seen at this status: 2026-10-12·8d ago
Feature Views let teams define ML features declaratively in Unity Catalog from Delta tables, Kafka streams or request-time data, using computations such as time-windowed aggregations, and materialize them to offline stores for reuse or to online stores for serving. The same definitions serve model training, batch inference and real-time serving. The APIs behind them were first released in Beta as the Declarative Feature Engineering APIs.
Status history
- GA
- Beta2026-03-09 · Mar 2026Release notes
Sources (3)
- 2026-10-12Release notes GA
- 2026-07-10Databricks Blog
- 2026-03-09Release notes Beta
Recent coverage
- 2026-10-14Run Live Energy Trading, Risk, and Analytics in Databricks video
- 2026-10-13v0.188.0 release
- 2026-10-13UnityCatalog 0.7.0 release
- 2026-10-10v0.165.0 release
- 2026-10-10v0.151.0 release
- 2026-10-09v0.164.0 release
Status and dates are extracted from public sources only. We don't see the private AE-shared roadmap. If something looks wrong, tell us.