How do you implement automated data quality monitoring and remediation workflows?

Data quality automation continuously validates data integrity across systems. Quality dimensions: accuracy (data correctly represents the real-world entity), completeness (all required fields populated), consistency (same data agrees across systems), timeliness (data available within SLA), uniqueness (no unintended duplicates), and validity (data conforms to expected format and business rules). Implementation: deploy data quality rules using frameworks like Great Expectations, dbt tests, or Apache Griffin. Rule types: schema validation (data types, nullable constraints), statistical profiling (detect distribution shifts, outlier spikes), referential integrity (foreign key relationships hold), business rules (custom SQL assertions), and freshness checks (data updated within expected window). Monitoring workflow: scheduled quality checks run after data pipeline completion, results aggregated into a quality scorecard per dataset, violations classified by severity (critical: blocks downstream, warning: investigate, info: log only), and critical violations trigger automated remediation or human escalation. Remediation patterns: auto-fix (standardize formatting, fill defaults for known patterns), quarantine (isolate bad records for review, continue processing good records), and escalate (notify data steward for manual correction). Implementation: data quality as a stage in every pipeline, quality gates that prevent bad data from propagating downstream, data quality dashboards for visibility, and SLA tracking per dataset. Feedback loop: track root causes of quality issues, implement preventive rules at data ingestion points, and measure improvement over time.