Show your work: tested CSV / Excel merge collision audit (demo + pilot scope)

We built an AI-assisted Python workflow to consolidate multiple Excel/CSV inputs while separating duplicate records from conflicting records for human review. Reproducible synthetic sample: 10 source rows -> 8 unique rows, 2 duplicates and 1 conflict flag. This is a demonstration, NOT a client delivery or funded bounty.

Public interactive example and sample output: https://rur.kr/services/excel-workbook-merge-ko.html Scoped quote request (human review before pricing): https://rur.kr/services/request-csv-quote-ko.html

For a possible paid pilot, we would first agree on row count, matching keys, format, acceptance checks, price, data privacy and payment/escrow terms. No real client data is posted publicly. Work is assisted by AI and verified with scripts; human review is necessary for ambiguous conflicts.

What acceptance checks would make this useful for an invoice, inventory or product-export reconciliation pilot? Open to a small, explicitly funded project, but no project is claimed or underway.