Catalogue QA with before-and-after reconciliation by Jasper AddymanCatalogue QA with before-and-after reconciliation by Jasper Addyman

Catalogue QA with before-and-after reconciliation

Portfolio project
Jasper Addyman
A wrong price or stock count can survive a bulk catalogue update if the team checks only how many rows imported. This Sheetstead Studio demo shows how to compare a staging export against the exact changes approved for release.
The check reads three CSV files: the before export, the approved list of changes and the after export. It joins them by SKU, validates prices and stock counts, and reports the expected and observed values for each discrepancy. It flags unapproved additions, removals and edits as well as missing or mismatched approved changes.
On a fictional batch, the first pass matched two of four approved changes and flagged five exceptions. One approved stock change expected 15 units, but the export still showed 12. After corrections, all four approved changes matched and there were zero exceptions across eight expected products.
The handover includes the SKU-level exception report, a repeatable Python check and a correction recheck. This is an original demonstration with fictional data. It reads staging exports and does not change a live store.
The studio developed the checker with AI assistance and verified the results against known fixture values.

0 likes · 3 views · Nov 11, 2026

Read-only Python audit of staging catalogue exports. Found five SKU, price and stock discrepancies in a fictional batch, then verified the corrected export.

Tags