A reproducible benchmark that scores data controls against known failure scenarios with precision, recall, and ground truth. Custom approach achieved perfect recall; industry baselines missed injected drift. 37 passing tests, 10/10 gates. Enterprise Data Trust, Chapter 3.
python benchmark data-engineering databricks ground-truth data-quality drift-detection data-trust medallion-architecture control-effectiveness
-
Updated
Apr 24, 2026 - Python