A WORKING EXAMPLE / FILE PROCESSING
Two exports.
One checked report.
Repeated order IDs. A missing quantity. Two prices for the same order. Run this small tool to see how we keep usable records and make the exceptions visible.
Fictional sample data. This is a working demonstration built by OpSmith AI, not a customer case study. It runs on this page using the samples below. No uploads, account, AI service, or API calls are involved.
Before: two order exports
Each row represents one complete order. The amounts are fictional USD values; taxes, shipping, and refunds are outside this example.
Ready. Run the demo to process all eight sample rows.
After: a report you can check
Exceptions stay visible
Review means a person must confirm the source information. This tool does not guess which conflicting price is right.
What the tool checks
- Every row needs an order ID, item, positive whole-number quantity, and a nonnegative price with two decimal places.
- For an exact duplicate, keep the first occurrence and explain the excluded copy.
- If values conflict for the same order ID, hold every version for review. Invalid rows also stay out of the total.
- Calculate in whole cents and account for every input row. The downloadable JSON contains the decisions and source row numbers.
This example handles its documented CSV format. A customer project starts with your sample files, your rules, and tests against the result you need.
Have a file task like this? ↗