AI Machine Learning Fraud Detection

AI Machine Learning Fraud Detection

Banking

THE CHALLENGE.

Even with lower check-processing times due to electronic payments and automated clearinghouse (ACH) transactions, banks must still manually verify millions of handwritten checks. Annually, banks risk losing millions as a result of check fraud by counterfeiters. Because a percentage of the funds is made readily available to the depositors, it’s critical to identify counterfeit checks quickly. To reduce the incidence of check fraud, a major bank partnered with Pegasus One to build a solution based on artificial intelligence (AI) machine learning to speed up check verification and lower costs.

Google TensorFlow™

AI Machine Learning Fraud Detection

Banking

THE CHALLENGE.

Even with lower check-processing times due to electronic payments and automated clearinghouse (ACH) transactions, banks must still manually verify millions of handwritten checks. Annually, banks risk losing millions as a result of check fraud by counterfeiters. Because a percentage of the funds is made readily available to the depositors, it’s critical to identify counterfeit checks quickly. To reduce the incidence of check fraud, a major bank partnered with Pegasus One to build a solution based on artificial intelligence (AI) machine learning to speed up check verification and lower costs.

Google TensorFlow™

APPROACH.

To meet the bank’s goals, our solution needed to identify fraudulent checks in real time, as well as reduce the number of checks requiring manual review. The bank already uses optical character recognition (OCR) and deep learning technology to scan checks, process data and verify signatures. Our model, based on Google TensorFlow™, uses a neural network to parse a historical database of previously scanned checks, including those known to be fraudulent.

We trained the neural network to use a set of comparative algorithms to distinguish good checks from anomalous ones. By automatically comparing various factors on scans of deposited checks to those in the database, our model flags potential counterfeits in real time. It assigns a confidence score to each scanned check, flagging it as good, fraudulent or needing further review. Our solution is scalable and configurable to the client’s evolving needs.

RESULTS.

Fraud prevention

48% reduction in fraudulent transactions

Higher Savings

$10M annual savings on fraud losses

FASTER RESPONSE

Response time less than 75 milliseconds, with up to 1,000 checks per second processed

VALUE.

Counterfeiters constantly develop new techniques to perpetrate fraud in financial services. Our AI solution operates with near-human intelligence to counteract the counterfeiters and reduce losses. Every transaction the model processes increases its accuracy of detection and adds to its enormous repository of historical information, so it’s continually learning the practices of habitual fraudsters to defeat them.