A small CSV cleanup.
A clear record
of every change.

Open a sample delivery.

Three invented products, two changed cells. Compare the original CSV, formatted CSV and cell-by-cell change log before you order.

Synthetic example, not customer work. This sample passed automated checks; real orders also require the full quality checks and AI review. Import SKU and price columns as text to keep leading zeroes and decimal formatting in your spreadsheet.

Before you place the order

• One UTF-8 CSV, with 1–100 data records plus its header row, 1–20 columns and no more than 20,000 bytes for the whole file.
• Use unique, nonempty column names without surrounding spaces or spreadsheet-formula prefixes. Every data record must contain the same number of fields as the header.
• Identify the columns that contain product identifiers and prices. Those values stay unchanged; this service does not repair identifiers or decide which prices are correct.
• Send a short, clear request. If your file contains unusually long, heavily escaped or complex text, ask through the marketplace first so we can confirm it fits the complete-review limit.
• Confirm that you have permission to share this non-personal, non-confidential file and permit AI-assisted processing. If you cannot give that permission, this offer is not a fit.

After checking these points, use this brief in the marketplace requirements and replace the bracketed parts. Do not paste private records into the brief.

File: [one CSV filename]
Identifier columns: [exact column names]
Price columns: [exact column names, or none]
Requested formatting: Trim surrounding spaces in ordinary text and apply formula protection where needed.
Please preserve every row, column header, identifier, price and product fact.
I have permission to share this file. It contains no personal or confidential data, and I permit AI-assisted processing.

Only include statements that are true for your file. The final scope and delivery deadline are confirmed in the marketplace order. A mismatch may require clarification there; this checklist is not automatic order acceptance.

See exactly what changed.

Three source products, five output columns, seven recorded changes. These files are actual outputs from our delivery tool, checked against the source worksheet. This is a synthetic example, not customer work.

Across all three products: one name is trimmed and six color/size values are split. Every row, identifier and price is retained.

Includes the source XLSX, mapped XLSX and CSV, exact rules, column mapping, seven-change log and quality-check report. The exception sheet is empty for this example.

Identifiers and prices are stored as text in the workbook. Choose text columns when importing CSV to preserve leading zeroes and decimal formatting. This example follows the supplied rules; destination acceptance is not guaranteed.

Follow every row back to its PDF page.

Two source pages, four products. These are actual outputs from our delivery tool, checked against the original PDF. This is a synthetic example, not customer work.

Includes the original PDF, Excel workbook, CSV, field-to-page evidence, page coverage and quality-check report. The exception sheet is empty because all four sample rows passed review.

The workbook stores identifiers and prices as text. When importing the CSV, select text columns to preserve leading zeroes and decimal formatting. This simple ruled-table example does not demonstrate OCR or every PDF layout; ask through Upwork if your layout is uncertain.