01 / Data quality

Clean records.
Every change, explained.

A small customer-data cleanup with a traceable audit log and a separate review queue.

Customer cleanup

Edit the source CSV

Required columns: record_id, customer_name, signup_date, city. Up to 1,000 rows. The selected date convention applies to slash dates.

Source rows
Cleaned rows
Exact duplicates removed
Items to review

Review before using the data

No guessed replacements

Audit trail

Original → cleaned → reason

Preserve identity

Names keep their case and accents. Blank values stay blank. Duplicate IDs with different records are retained.

Export safely

CSV cells starting with formula characters are prefixed with an apostrophe so spreadsheets treat them as text.

A bounded prototype

This demo uses four columns and explicit rules. A client project starts by agreeing on its own schema and validation rules.