Digital Ops and Check Fraud – Prevent Operational Losses

Rising financial crime continues to pose a significant risk to banks and financial institutions. Fraudulent checks routinely flood the system and lead to millions of dollars in operational losses for banks. In addition, the lack of robust fraud detection processes and intelligent systems impacts profitability and customer experience.

In this case, the bank faced high operational losses of over USD 45 million per annum due to deposit and check fraud. They had an analytics engine and detection system in place alongside an onsite team of 150 people for manual scrutiny, but they could not cut down on the losses. This inefficiency stemmed from poorly framed and documented processes, lack of adherence to the existing process, and inadequate checks and balances programmed into the analytics engine. The analytics engine failed to detect fraudulent checks worth USD 25 million per annum. For those cases that it did detect, errors and omissions from the manual team led to an additional USD 20 million in losses.

The client wanted a partner with strong experience in check fraud detection who could help them establish a robust process, minimize the need for manual intervention, and cut down operational losses to less than USD 20,000 per day. At the same time, to manage customer experience, the requirement was that no more than 10% of the checks coming in could be held for scrutiny on any given day. In addition, 98% of the checks held back must be for legitimate concerns allowing only a 2% margin of error. These were stringent SLAs, but the SLK team was more than prepared to meet them.

SLK came on board and took over the entire process offering the client a 70% discount on the existing processing costs. The first step was to fix the process itself and stabilize and strengthen it. We drew up a new procedure with stringent checks and balances. We moved the team offshore and trained them on the improved process, ensuring adherence. Once that was in place, we deployed intelligent automation to cut down manual intervention.

We tackled the downstream process first, deploying UiPath to automate repetitive tasks and build manually proven checks and balances into the system. Next, we deployed artificial intelligence on the upstream process to cut down the number of checks coming in for review. The AI engine improved detection by checking for significantly more parameters than the original analytics engine. Check stock identification, verification, and risk scoring improved the accuracy of detection. Increased automation while reducing operational losses also reduced the dependence on the manual team freeing up employee bandwidth.