Data Quality Management
Define column contracts, quarantine failing records and record corrections against an unaltered source dataset.
Application guideVerifying your session…
Using Data Quality Management
Quarantine bad records and resolve their issues against a reproducible dataset contract.
Column contracts · Record quarantine · Traceable repair
- Inputs and working data
- Original CSV and SHA-256, typed column rules, unique sets, field-level corrections and review reasons.
- Working in the editor
- Contract editor, paged quarantine queue, record inspection and repair history.
- What can I produce?
- ZIP of original and accepted CSV, exact structured accepted/quarantined records and provenance manifest.
- What are the limits?
- 5,000 records, 60 columns, 5,000 tracked field repairs and a 1.5 MB source import.
- Example
- Quarantine a negative count, record the correction and verify original source, clean output and reason in the release package.