While digital transactions have increased at a fast pace, checks still account for nearly a quarter of the total number of financial transactions in the US and with check fraud becoming increasingly sophisticated, organizations are actively seeking to track and prevent it.
Today most banks employ automated means as their primary defense to fight check fraud. Additionally, banks also have teams of human agents doing manual verification of checks. While human experts can do a good job, often manual verification is time-consuming and error prone. With increasing volumes and costs of processing, there is a strong need to optimize the process of manual verification of checks to improve productivity and accuracy as well.
The Infosys Check Fraud Detection is a powerful ‘human-in-the-loop’ solution for the manual check verification process. It can deliver productivity savings of up to 50% and also ensure higher accuracy. The solution uses a powerful combination of artificial intelligence and machine learning (AI/ML), computer vision, analytics and other approaches to detect check anomalies not apparent to the human eye. The solution also has a smart UI which provides a 360o view of the customer’s banking activity enabling a comprehensive audit in a shorter timeframe.
Productivity savings: Faster turn-around time and Reduced cost of Check Validation (~40-50% savings in effort during high-volume lab-based simulation), Prioritized Reviews: Rating of the alerts to highlight more likely fraud cases (~85-98% accuracy for various aspects like signature, amount, logo etc.) to improve the effectiveness of these checks.
Infosys Check Fraud Detection addresses different aspects of the fraud detection process using state-of-the-art AI techniques to deliver productivity savings of ~50%. The solution uses advanced algorithms to verify the authenticity of the check to cover various parameters including payer, amount, date, anomalies in the signature, text, character spacing, logo, etc. Further, comparison is also done with prior returned (fraud checks) from the same bank. Data analytics is used to verify customer details including the age of the account, transaction history, and patterns to determine potential red flags.
The solution has a unified UI which brings together all elements required by check verification by the agents covering check images (current and past), audit checklists with rules enabled, as well as relevant banking data of the customer. The 360o enables high productivity for the agents which doing their verification. Finally, the solution includes a maker-checker function that is useful for additional verification by a supervisor, for instance, in cases involving high-value checks or complex scenarios.
Typical approaches rely on a standard operating procedure with a fixed set of parameters to validate manually. However, with an AI approach, the Infosys solution can consider a wider range of characteristics of the check and identify new varieties of outliers.
The solution is built on Infosys’ deep expertise in the fraud space coupled with extensive experience of running client operations in check fraud and related areas. It has also leveraged Infosys’ expertise in AI/ML built over the past two decades and is built on the foundation of Infosys Topaz.
AI-driven automation identifies all possible anomalies enabling agents to make quicker and better decisions
Solution automates the analyses of various parameters to bring more rigor to the validation process and prevent mistakes. Maker-checker workflow enables quick routing of high-value / complicated cases to the supervisor
360o UI ensemble seamlessly integrates with client systems to bring relevant check images, data etc. into a single ecosystem for quick and accurate decision-making