Data Quality Agent

Profile data, propose rules, simulate impact, detect incidents and recommend actions.

Purpose

The Data Quality Agent keeps every published data product trustworthy. It profiles data across the 6 DQ dimensions, proposes plain-English + SQL rules with severity and rationale, simulates impact before deployment, monitors every refresh, and opens explainable incidents with recommended remediation — all reviewed and approved by a human steward.

Solves

Silent data breaks, missed SLAs, noisy alerts, unexplained anomalies, slow root-cause analysis.

Delivers

DQ scorecard, approved rule set, incident timeline with lineage, recommended remediation, SLA metrics.

Handoff

Incidents route to stewards; approved fixes feed back to improve rule proposals over time.

1Select Data Product

2Profile Quality

3Propose Rules

4Simulate Rules

5Detect Incident

6Review Action

Step 1 · Select Data Product

Data Product

Source

Layer

Criticality

Refresh SLA

Data Steward

Operational SLAs (last 30 days)

Live monitoring posture for this data product.

Mean time to detect

4 min

-38% vs prior

Mean time to resolve

1 h 22 m

-21% vs prior

Incidents open

+1 vs prior

SLA compliance (30d)

97.4%

+0.6% vs prior