OrboGraph: Securing Payments with Image Analysis, AI, and Automation

OrboGraph: Securing Payments with Image Analysis, AI, and Automation

Follow OrboGraph on :

Joe Gregory, Vice President of Marketing, OrboGraphJoe Gregory, Vice President of Marketing
Banks and other financial institutions (FI) are investing in transformative technologies to deliver a superior customer experience using mobile as well as real-time payments. However, FI’s must also continue to be diligent on legacy payments where fraud losses continue to persist due to traditional systems and processes as well as inadequate cybersecurity measures. The check is one such legacy payment.

While analytics-based solutions help companies detect a variety of fraud scenarios, the systems also lead to false positives and the unnecessary overhead of a large database of suspicious accounts and transactions when it comes to check payments. These existing security gaps give impetus to increased fraudulent activities such as check forgery, counterfeiting, alterations, stolen check deposits, and identity theft. Additionally, error-prone manual processes make it difficult for FI’s to monitor and review on-us and deposited check transactions without analyzing the image itself.

Serving the financial services industry since 1996, OrboGraph is delivering intelligent payment automation platform for check processing. Its flagship product OrboAnywhere helps banks take a security and compliance-driven approach to enabling real-time fraud prevention on check transactions. “Our platform combines cutting edge analytics, self-learning, image analysis, AI, and deep learning technologies to close the security gaps for corporate clients and financial institutions,” says Joe Gregory, Vice President of Marketing at OrboGraph. “The world is moving to real-time, and we enable FI’s to prevent fraud in real-time through any self-service or omnichannel deposit location.”

OrboAnywhere is comprised of several modules relevant for fraud and risk mitigation. Anywhere Fraud identifies counterfeits, forgeries, and alterations, analyzing both check images and transaction data for each customer account.

Our platform combines cutting edge analytics, self-learning, image analysis, AI, and deep learning technologies to close the security gaps for corporate clients and the financial institutions

The system blends self-learning algorithms to develop a rich, behavioral profile for each account with image analysis tests including check stock validation (CSV) and automated signature verification (ASV). These tests provide scores which are used for identifying potential fraudulent transactions.

Banks that adopt image analysis technology can reduce fraud risk and operational overhead. As an implementation highlight, a client integrated OrboGraph’s image analysis with its legacy analytics system, resulting in a reduction of fraud loss by nearly $3M on a yearly basis. In another instance, a large bank implemented an end-to-end suspect review platform which resulted in reducing fraud defects to only 3 per million items processed…better than Six Sigma!

Another module of interest for fraud prevention is the Anywhere Validate module. This component utilizes advanced AI recognition and scoring technology to mitigate the risk of payment negotiability. For example, checks with missing signatures, missing endorsements, or post/pre-dated checks can be identified automatically in a check deposit workflow such as Mobile RDC or ATM deposits.

OrboGraph follows a well-defined client onboarding process by using either historical data or real-time transactional data. If done on a batch basis using historical data, the data collected over the last 90-120 days is fed into the platform to train the system. In the case where real-time transactional data is used, the self-learning algorithms in OrboAnywhere take 90 to 120 days to learn from the incoming data before going live.

OrboGraph is collaborating with various financial institutions to develop more AI-centric solutions that are aligned to their specific business needs. Also, the company is expanding its capabilities in image recognition, AI, and deep learning models to help clients and business partners increase the scope of secure automation in the payments arena, particularly in the area of fraud prevention and compliance.