For data owners and leadership: is our data trustworthy, where is the risk, and is it improving?
RAG band: Green ≥98% · Amber 95-98% · Red <95%
| Dataset | Domain | Weakest dimension | Score | Business impact | Status | Owner |
|---|---|---|---|---|---|---|
| Reference codes | Reference Data | Validity | 92.0% | Mis-routing, report breaks | RED | D. Steward |
| Transactions | Sales | Timeliness | 94.5% | Reports lag reality | RED | C. Owner |
| Customer master | Customer | Accuracy | 95.8% | Failed contacts | AMBER | A. Steward |
| Accounts | Finance | — | 99.1% | Healthy | GREEN | B. Owner |
For data engineers and stewards: which checks are failing, on which tables, and is anything past its freshness SLA or blocking the pipeline?
| Check ID | Target | Rule type | Failed rows | Threshold | Runs in | Status |
|---|---|---|---|---|---|---|
| TR-004 | reference.country_code | accepted_values | 1,920 | ≥99% | dbt | RED |
| TR-008 | fct_transaction.updated_at | freshness | late 3h | within SLA | dbt source freshness | RED |
| TR-003 | dim_customer.email | regex | 760 | ≥98% | DQ engine (Python) | AMBER |
| TR-006 | fct_orders.amount | range | 280 | 100% | dbt | AMBER |
| TR-001 | dim_customer.customer_id | uniqueness | 0 | 100% | dbt | GREEN |
Illustrative example dashboard. Figures are fabricated to show layout and method, not real client data. Built by Bernard Millet.