Data Mining for Fraud Detection

Procurement Integrity Efficiency Review

PIER Protection

GFAS is focused on enhancing and improving the integrity and accountability within organizations. We specialize in developing unique approaches to enhance profitability by eliminating fraud, waste and abuse.

At GFAS, our team combines deep experience with cutting edge techniques to provide you with complete confidence in the integrity and efficiency of your supply chain. Our utilization of proprietary software, coupled with the unparalleled experience of seasoned professionals, has enabled us to create a comprehensive suite of machine-based learning and real time algorithms. These tools are designed to detect patterns, in addition to highlighting risks associated with procurement fraud and inefficiency.

Utilizing a SAAS based approach, we have developed a phased approach that enables us to analyze and normalize large amounts of disparate datasets, while performing unique tests that highlight unusual patterns and attributes identifiable with inefficient and potentially fraudulent procurement activities.

Not only are we able to assess and analyze the accounts payable cycle, we also apply these powerful tools to review other areas of your organization such as payroll, grant accounting, disbursements and procurement card activities.

Combining specific sets of algorithms, we develop rules that are unique to your organization’s procurement and payment “behavior” As the datasets grow, our program learns and adopts rules to adapt to evolving procurement patterns, confirming outliers and exceptions while minimizing false or misleading positives.

Accounts Payable

Our A/P (accounts payable) tool will help to uncover overpayment and recovery opportunities such as missed discounts, inconsistent terms, pricing errors and receiving anomalies.

Using ten different axes, the tool also identifies duplicate payments and vendors while cross matching addresses, zip codes, tax IDs, bank routing numbers and other specified identifiers.

Our program uncovers potentially fictitious and suspicious vendors by combining attributes such as sic codes, DUNS information, GPS, mail drop addresses, unit price and receipts data. The A/P tool also highlights certain behavioral patterns such as weekend payment dates and PO Boxes, while cross matching to the employee payroll master file by zip code and various address attributes.

• The application of a three-way match
• Prohibition of product substitutions
• Prohibition of violations of approval ceilings
• Detection of splitting or changes in volumes
• Detection of sequential numbering of invoices and/or rounding of amounts
• Detection of all aspects of suspicious or inefficient procurement

Procurement Card Activity and Controlled Spend Monitoring

The PIER program applies powerful analytics to review company credit card or procurement card purchasing patterns, effectively identifying and strengthening internal controls. Our system can point to unauthorized use of a credit card in prohibited SIC codes and excessive authorized amounts, by cross matching purchases and credits against the purchasers’ home addresses.

Grant Accounting for Educational Institutions

Our unique module utilized for reviewing grants is built on the “yellow book” standards and provides necessary metrics to the post-award process. This module reviews the sufficiency of interface with a grant and the general ledger package, highlighting the audit readiness standards of any grant. It detects personnel costs applied without documentation, reported costs not reconciling to the accounting records, “in kind” contributions provided without support, and improper support for cost allocations. Critically important, it will detect a failure to utilize and allocate grant amounts within proper time limits.

Payroll Management

Payroll is a crucial fuel that keeps the engine of your organization running. It also represents a potential risk that is often overlooked. The 2020 Report to the Nations on Occupational Fraud and Abuse published by the Association of Certified Fraud Examiners (ACFE), reported that payroll fraud may be not as uncommon as one might think. The ACFE survey highlighted nine sub-categories of asset misappropriation schemes, each representing a specific method used by employees to misappropriate organizational resources. Payroll frauds accounted for 9% of the total asset misappropriation category in frequency. The median loss was $ 62,000.00.

The identification of payroll fraud is uniquely challenging due to a combination of risk factors, which make the payroll cycle a rich target environment for fraud. According to the ACFE, payroll fraud occurs in 27 percent of all businesses. Small organizations with less than 100 employees are nearly twice as likely to experience payroll fraud than larger organizations.

Our payroll PIER program utilizes several disparate data sets including the Social Security Death Master Index, GPS address matching and the gross versus net algorithms to point out anomalies, bank deposit consistency matching, length of service matching, non-repeating salaries, one time pay set ups, dormant and re-resurrected employee files in addition to other analyses.

Utilizing these tools and our experience in the payables, payroll and accounting cycles, we bring a common approach to deployment that enables us to:

• Normalize large sets of disparate data
• Enhance and measure data and meta data quality for analysis
• Easily adapt and deploy rule sets and enable “machine learning” to refine output
• Provide forensic data analysis to help guarantee data quality
• Perform associative linking that helps to provide semantic cluster analysis and keyword identification
• Provide unparalleled anomaly detection to reduce manual analysis while continually improving accuracy

Our experts can summarize and prioritize in an easily understood output, any suspected transactions, vendors or patterns. We subsequently rank them by likelihood of potential risk to the organization, based on the results of the analyses run. This process will assist in determining future action or audit tasks and drastically reduce non-productive time spent sifting through irrelevant datasets.
The key to the success of these tools? Combining cutting edge technology engines with seasoned and superior fraud examination professionals. The exceptions found warrant further investigation. When you hire individuals with the experience, discretion and professionalism of the GFAS team, this greatly enhances the outcome of the process. Determining whether these outliers can be explained or require a trained forensic approach to a successful resolution is accomplished much more quickly.