SchemaSentry

SchemaSentry monitors the “shape” of your data across sources and pipeline stages and alerts you when something changes in a way that will break downstream jobs or dashboards. Instead of generic data quality rules, it focuses on high-signal schema drift: column additions/removals, type changes, nullability shifts, enum/cardinality explosions, and contract violations between producers and consumers. It plugs into common warehouses and lakehouses, snapshots metadata on a schedule, and compares changes against approved baselines. When drift is detected, it opens a ticket/PR with a suggested fix (e.g., update dbt models, adjust casting, add backfill), plus a blast-radius report showing which tables, models, and dashboards are impacted. This is not magic: it won’t fix bad modeling or messy ownership, but it will reduce surprise breakages and the endless “why did this dashboard die?” fire drills.