SLSepiaLog
Deliver · SepiaLog feature

Data quality and sensitive-data checks before problems travel downstream

Compare the current data with its previous state and screen locally for quality changes and potential personal-data indicators.

What can an automated research data health check detect?

A data health check can flag structural changes such as row loss, duplicate identifiers, new missing values, column changes, type changes and implausible ranges. A separate sensitive-data screen can identify potential personal-data patterns for researcher review.

The research problem

A small unexpected change can travel through an analysis unnoticed. Sensitive fields can also remain in a sharing branch because risk review happens only at the final upload.

Research actionConnected documentationRecoverable output

How the workflow works

  1. Run the local quality comparison against the last saved state.
  2. Review alerts, warnings and informational changes.
  3. Run the sensitive-data screen before selecting access or preparing a public branch.

What SepiaLog provides

  • Row, column, type and missing-value changes
  • Duplicate IDs and implausible ranges
  • Potential names, emails, addresses, dates, identifiers and precise locations
  • Suggested access level and anonymization checklist

Common questions

Does the GDPR screen certify compliance?

No. It highlights potential risks for review and cannot replace legal or institutional guidance.

Does scanning upload the data?

The checks are designed to run locally over project files.

What should happen after an alert?

Inspect the affected records, confirm whether the change is expected and document the decision.