Our Experience
Project: The World Bank
Created and implemented vendor file maintenance directive for The World Bank in conjunction with Phillip Rogers.* Identified duplicate vendor records using fuzzy (non-exact) matching on name and address, including many foreign names that are not as standardized (and thus more difficult) as U.S. data. Additionally, grouped vendors into "vendor families" using proprietary link analysis algorithm.
Project: National Football League (NFL), Methode Electronics
Provided data management support for 2 major Accounts Payable (A/P) recovery audits, one for the NFL and one for Methode Electronics; both projects were performed in conjunction with Phillip Rogers.* The activities performed during the A/P audits:
- duplicate payment identification
- duplicate employees
- duplicate vendors
- potential ghost or fictitious vendors
- situations where a vendor and employee shared a common address
- payment anomalies such as above-average payments and rounded-amount invoices
- unusual payment behavior via Benford's Law analysis
- check register / A/P file discrepancies
- potential check fraud
- rapid invoice volume increases
*Phillip Rogers is a licensed Attorney in the State of Texas. The referenced projects above, as well as others, were completed before Mr. Rogers joined MAXIMUSalliance. Mr. Rogers is currently a principal and acting CEO of MAXIMUS, a national firm with local state and local tax expertise. MAXIMUS provides State and Local Tax (SALT) solutions and Business Incentives to companies with multi-state representation. For Automated Auditor, LLC referrals please visit
www.maximusalliance.com to obtain Mr. Roger’s contact information.
Project: Bruce Dubinsky, MST, CPA, CVA, CFE
Created a suite of name-matching algorithms for the purpose of identifying potential fraudulent payments to known terrorists, as identified by the Department of Homeland Security. The programs searched for known terrorist names among millions of payment transactions using a technique called "fuzzy matching". Fuzzy matching strengthens a search query by taking out vowels and considering consonants such as "c" and "k" alike. For example, fuzzy-matching allows the names "Mohamed Al Rashka" to be considered similar to "Mohommad el Rashca". Many corporations and government entities have to perform similar name-scrubbing queries to maintain compliance with the USA Patriot Act.
Client: Bruce Dubinsky, MST, CPA, CVA, CFE
One of Mr. Dubinsky's clients (confidential) was engaged in a lawsuit with a large accounting firm regarding a phony tax-shelter scheme. Mr. Dubinsky hired Automated Auditors to provide data management and statistical analysis support to identify the odds of making a profit using the faulty tax-shelter scheme. The process involved importing large amounts of stock data into SAS datasets and performing a Monte Carlo study to determine the probability of making a profit off of the tax-shelter scheme. Using binomial probability theory and logistic regression, Automated Auditors was able to conclude that the probability of turning a profit on the tax shelter scam was zero.
Project: Confidential
Utilized the Social Security Administration's (SSA) Death Master File (DMF) to identify potential healthcare claims made to deceased insurance recipients. Identified millions of dollars paid to healthcare providers that were potential overpayments made for services rendered after the patient's death.
Project: Various
Developed SAS algorithms to identify healthcare fraud, waste, and abuse, in both the private and public sector. Some of the algorithms include, but were not limited to:
- Unbundling
- Procedure Upcoding
- DRG Upcoding
- Services Not Rendered
- Rent-A-Patient Scheme Identification
- Duplicate Claim Identification
- Over-Utilization
- Physicians Working Many Weekends/Holidays and Consecutive Days
- Ambulance Trips to Nowhere (no inpatient admission or ER visit)
