Supplier formatter for the ERP · v1.0 · 2026-07-19
Drag your supplier file here (Excel or CSV): it never leaves this page.
Expected columns: Reference ; Description ; Delivery date ; Weight (kg) ; Quantity
The sample file has three defects planted on purpose; you can also to test the full flow.
File read:
rows read
transformed
anomalies
Anomalies: rows excluded from the corrected file, never silently fixed
Before: as received
After: ready for the ERP (changes highlighted)
Your original is untouched; the output is named
What this tool does, exactly:
- It reads an Excel (.xlsx, .xls) or CSV file and locates the header row by itself (Reference, Description, Delivery date, Weight, Quantity columns; the exact spelling may vary).
- References: uppercased, spaces removed. A duplicate reference is reported and the row is excluded.
- Dates: converted to the YYYY-MM-DD format the ERP expects. Accepted: real Excel dates, day-first DD/MM/YYYY (also with dots or dashes) and YYYY-MM-DD. An ambiguous or impossible date (e.g. US format MM-DD-YYYY) is reported, never guessed.
- Weights: a decimal comma becomes a point; a trailing "kg" suffix is removed. The weight is assumed to be in kilograms; no unit conversion is done. A missing, unreadable, zero or negative weight is reported.
- Quantities: must be positive whole numbers, otherwise the row is reported.
- All five columns are mandatory: any missing value (including the description) is reported and the row is excluded from the corrected file. Nothing is silently fixed.
- The produced file is an ERP-ready Excel: dates as YYYY-MM-DD text, weights and quantities as numbers. It bears a different name from the original (e.g. supplier_formatted_2026-07-19.xlsx); your original file is never modified.
- These rules are starting assumptions: if one of them does not match your reality (weight unit, optional columns…), the tool can be adjusted: send it back to the AI conversation that produced it.
This tool runs entirely in your browser: no data is sent over the internet.