AI-Powered Disability Claims Handling Qorus Innovation in Insurance Awards 2018 - Winner

The solution applies a range of different AI technologies to automate a process that traditional automation technology was unable to automate, resulting in most customers receiving their disability payment within days instead of several months. By segmenting all incoming claims based on risk and complexity, the solution also ensures that only the cases that truly require human involvement are processed manually, highly alleviating the claims management department. Combining business rules with machine learning, the solution first segment all incoming claims in three segments – green, yellow and red – based on their risk and complexity. Green cases are sent to straight-through-processing using RPA. For the yellow cases, requiring some further assessment, the solution uses computer vision to read pdf documents containing medical information, and then text analytics to extract the pieces of information that are required to complete the assessment, before the case is sent to an RPA robot for finalization. This way, the case handlers are able to focus on the red cases – claims with high risk or complexity that really need manual handling. All cases also go through a machine learning-based fraud detection engine flagging cases that might have a high fraud risk.