Microsoft Fabric's Materialized Lake Views (MLVs) reached general availability in March 2026, collapsing the traditional medallion architecture into declarative SQL. Previously, building bronze-to-silver-to-gold pipelines required stitching together notebooks, pipelines, schedules, custom validation code, and Monitor Hub. Now a single SELECT statement handles transformation, storage, refresh scheduling, and dependency tracking automatically.

Each MLV moves through Create, Refresh, Query, and Monitor stages. During refresh, Fabric chooses optimal strategies: skip unchanged data, process only changes incrementally, or rebuild entirely. Incremental refresh requires Delta change data feed (CDF) enabled on source tables and append-only workloads. Without CDF, systems fall back to full refresh. The materialized results store as Delta tables, making them queryable by Power BI Direct Lake, Spark notebooks, and SQL endpoints without special handling.

The GA release addressed preview limitations with five key improvements. Multi-schedule support allows different refresh frequencies within a single lakehouse—finance hourly, analytics every six hours. Broader incremental refresh coverage expanded supported SQL constructs significantly. PySpark authoring, in-place updates with REPLACE, and stronger data quality controls round out the production-ready features.

MLVs eliminate the debugging nightmare of layered orchestration failures, where a bronze notebook error at 3 AM breaks downstream silver layers silently. For data engineers managing complex lakehouse architectures, this represents genuine operational simplification rather than just another abstraction layer.