Data Quality Incident Detection Blind Spot
warningOrganizations experiencing data quality incidents (bad data reaching dashboards, ML models) that DataHub's observability should catch, but failures occur because assertions are not configured or monitored on critical datasets.
Track business-reported data quality incidents (from tickets, incident management systems, user reports) that were not detected by DataHub assertions. Monitor coverage of critical datasets with active freshness, volume, schema, and custom quality checks. Alert when high-impact datasets lack quality monitoring or when incidents occur on unmonitored assets.
Implement proactive Data Quality coverage tracking: catalog critical datasets based on usage (query frequency, dashboard dependencies) and business impact (revenue reports, ML features). Bulk-create Smart Assertions for volume and freshness on all high-priority datasets. For incidents that occurred on monitored datasets, review assertion configuration - tune sensitivity, add column-level checks, implement custom SQL assertions. Use DataHub incidents API to raise incidents when quality checks fail and notify dataset owners via Slack. Establish SLO targets for quality check coverage across critical data domains.