He realizes in one of the charts, that something was familiar. He thinks and takes another look at his results, and uses IDEA’s Visualize and Discover tabs for inspiration. Did he decide to compare Active Directory vs Employees, and maybe then Social Security Numbers and addresses? Holy smokes! He found a Ghost Employee! Who had the access and control of the system to be able to add both a Phantom Vendor and Ghost Employee? Using Boolean Search tactics and IDEA’s built-in text search he further checked the data. The bulk of the amounts are under 1 million, but the quartiles revealed some entries in the 2-million-dollar range. Using Python inside IDEA Jack queried his data, and used a Five Number Summary and the Numpy Library to calculate quartiles to find his outliers. Who and How? He had to find those answers.
He was able to segment a population that needed further investigation. Maybe he should go one step further using Fuzzy Matching. He decided his next weapon against fraud may be needed. Sure enough, there were some duplicates, but none that were the direct cause of the anomaly. What is the easiest way to find anomalies? He decided that maybe Duplicate Detection could help. He pulls up his multi-tool, CaseWare IDEA and begins to think. He knows, as an auditor, this is a big deal. Jack goes back to his desk, wondering how, out of all the data before him, he could possibly find the mistake. His inspector looks up from his work and says, “There is a simple mistake. “Inspector, I have found a problem in the AP file,” he says.
0 kommentar(er)
