# Warehouse Cleanup Change Summary

Public synthetic sample for Rowva artifact inspection.

## Source

- File: `warehouse_messy_data.csv`
- Source rows: 18
- Operational export rows: 15
- Cell-level fixes logged: 43
- Rows excluded from operational export: 3
- Conflicts preserved for review: 4

## Automatic fixes

- Normalized SKU values to `sku_####` format.
- Converted mixed dates to ISO `YYYY-MM-DD` where unambiguous.
- Converted currency strings to numeric unit costs.
- Canonicalized warehouse names such as `NY whse`, `New York warehouse`, and `NYC`.
- Trimmed whitespace from quantity fields.

## Rows excluded from the clean export

- Row 2: duplicate of row 1.
- Row 6: subtotal/summary row, not a product record.
- Row 12: duplicate of row 11.

## Review items kept

- Rows 4 and 5 disagree on supplier for `sku_8811`.
- Rows 7 and 8 disagree on usable quantity for `sku_4030`.
- Rows 9 and 10 look like possible duplicates for `sku_7710`.
- Rows 14 and 15 contain an ambiguous date disagreement for `sku_4404`.

## Readiness

- Operational Clean Export: ready for internal review and spreadsheet use.
- Compliance-Safe Export: ready after direct contact details are masked.
- Reporting Package: partial because unresolved supplier/duplicate conflicts remain.
- AI Package: blocked because direct contact details and review conflicts remain.
- Audit Trail Bundle: available.