FLIGHT MANUAL — DEEP DIVE

This project engineered a full-population SQL sweep across the enterprise ledgers of a multinational client, replacing the traditional sample-based approach with a comprehensive interrogation of every transaction.

The objective was to detect duplicate payments and journal-entry anomalies that sampling might miss. By querying the entire population rather than a subset, the audit provided complete assurance over the targeted assertions — a level of coverage that sampling cannot achieve.

The SQL queries were designed around specific risk patterns: invoices paid twice under slightly different reference numbers, journal entries posted at month-end with no supporting documentation, and payments to vendors whose bank details had been recently amended. Each query produced an exception list that was then individually investigated.

The result was a measurable reduction in leakage and a defensible audit opinion backed by full-population evidence rather than statistical projection. The approach has since been adopted as the default for high-risk assertion testing where data is available.