Slavia pojišťovna a.s. - case study

Requirements

In the past, out of 7,000 insured events (hereinafter referred to as IEs), approximately 190 IEs (2.7%) were reviewed annually, and a positive fraud indication corresponded to about 30% of those investigated (0.85%). The average amount of an IE was around CZK 67,000; and the total amount of IEs withheld was around CZK 4 million.

The initial indication of a suspicious IE was provided only by adjusters, there were no uniform rules for the detection of suspicious IEs, the context was not clearly visible in the system, the investigator processed only IEs indicated by adjusters.

The customer wanted to increase the efficiency of the identification of suspicious IEs and thus, of course, to increase both the efficiency of the investigators' and the adjusters' work and, last but not least, to reduce the expenditure on unauthorised IE payments.

Solution

The motivation for implementing the SAS Visual Analytics, SAS Visual Statistics was to increase the efficiency of capturing suspicious IEs with a high probability of fraud and to optimize the entire fraud detection process.

The planned estimate of the implemented rules was lower, but during the project they were already increased, which turned out to be necessary to deliver the contract of the quality expected by the customer. Thanks to the sharing of experience of both parties, mutual communication and a proactive approach during the implementation, unpleasant findings in the final testing phase were avoided.

At the same time, quite extensive modifications of the standard graphical interface (dashboards) were made to increase the comfort of working with the implemented tool according to the customer's requirements.

The consistency of the Fraud Management System data was ensured, among other things, by connecting to the customer's core insurance system so that it not only receives the data, but also returns information about the result of the investigation to the system.