Data quality
Checks completeness against the documented quality policy and raises a ticket for each violation, assigned to the right owner. For data stewards who need issues visible and owned.
Ask where quality needs attention and approve the tickets worth raising.
- Azure SQL databases (live, read-only)
- Confluence quality policy
- Jira tickets
- Collibra owners
- Findings per table against the policy
- Tickets for violations, assigned to owners
Four steps from start to finish.
- 1Ask
A steward asks which tables have data quality issues.
- 2Measure
The agent checks completeness on the live tables against the policy in Confluence.
- 3Check
It looks at Jira for open tickets and at Collibra for the responsible owner.
- 4Raise
For gaps without a ticket it proposes one. The steward reviews and approves it.
Quality issues are found by whoever notices them, and nobody is sure whether anyone is already working on them.
The agent queries the live tables, compares completeness with the policy written in the wiki and checks the board for open tickets. For a gap with no ticket it proposes one, with a summary, priority and the owner from the catalogue.
What it means for a team like yours.
Find tables with missing values in the retail data. Illustrated with fictional demo data, not client data.
Check invoice records for missing mandatory fields. Illustrated with fictional demo data, not client data.
Check that every deal has the records it needs across the trade tables. Illustrated with fictional demo data, not client data.