I will safely deduplicate and standardize your business CSV

 I will clean one authorised, nonsensitive business CSV using only exact rules you select and confirm.

 Supported:

 - remove exact duplicate rows or duplicates by columns you choose;

 - trim leading/trailing spaces;

 - apply your selected upper-, lower- or title-case rule;

 - standardise dates to your selected format;

 - standardise identified numbers;

 - make exact find-and-replace changes you list;

 - require selected columns or allowed values;

 - preserve text identifiers, including leading-zero product codes.

 You receive cleaned.csv; rejected.csv for rows failing confirmed rules; changes.csv as an audit trail; report.json with row

 accounting/checks; and a plain-language delivery note.

 This is fixed-rule processing, never fuzzy or inferred correction. I will not guess a value's meaning. Unsupported,

 contradictory, ambiguous or sensitive orders are declined or cancelled safely.

 Not supported: fuzzy matching or inferred correction; enrichment; merging/joining files; column renaming/reordering; new text;

 spreadsheet formulas/macros; databases; personal, medical, payment/authentication, credential, regulated or other sensitive

 data; unauthorised files.