Term Kind Topic What it is
Data Observability practice Data Observability Continuous automated monitoring of data health — freshness, volume, schema, distribution and lineage — designed to detect silent data failures.
Pipeline Anomaly Detection practice Data Observability Monitoring row counts, distributions, freshness and schema for unexplained change, because data pipelines fail silently far more often than they error.
Silent Data Failure No Error, Plausibly Wrong, Quiet Breakage concept Data Observability A data problem that produces no error - the pipeline succeeded, the schema was valid, the numbers were plausible - which is why error-based monitoring cannot detect it and why it persists for weeks.