Synthetic CSV reconciliation: exact IDs and control totals

Synthetic CSV reconciliation — exact IDs and checked totals

Fictional demonstration, not client work. AI-assisted preparation with arithmetic verification.

Two exports contain 7 and 6 rows. Left IDs/quantities: 001/10, 002/7, DUP/2, DUP/3, blank/4, LEFT/6, 0007/1. Right IDs/quantities: 001/8, 002/7, DUP/5, blank/9, RIGHT/11, 7/2.

Method: exact case-sensitive string IDs; integer quantities; differences are right minus left. Blank IDs are unresolved. If a key repeats in either file, every row for that key on both sides is quarantined pending an owner-approved matching rule. Leading zeros are meaningful.

Findings: 001 differs by -2; 002 matches. Left-only keys are LEFT (6) and 0007 (1); right-only keys are RIGHT (11) and 7 (2). No assumed match between 0007 and 7. DUP and blank-ID rows remain unresolved.

Control totals: left 33, right 42. Unresolved quantities: 9 and 14. Eligible quantities: 24 and 28. Matched quantities: 17 and 15. Unmatched quantities: 7 and 13. Every quantity is accounted for: 33 - 9 - 2 - 7 + 13 + 14 = 42.

The report identifies differences; it does not decide which source is correct or overwrite inputs. The fixture was checked with arithmetic assertions and a second calculation. This demonstrates matching rules and reporting, not production throughput or business results.