Early detection of insurance fraud
Infinity Property & Casualty Corp. (now owned by Kemper Corporation)
The objective for the first phase was to implement a batch scoring solution, which would use a combination of business rules and predictive modeling to identify potentially fraudulent claims that would need to be investigated further, and identify closed claims with potential for subrogation.
At the conclusion of the first phase, the leadership at Infinity Property & Casualty were sufficiently comfortable with the solution and the benefits that it brought that a second phase was scheduled to extend the solution to scoring claims in real-time during the first notice of loss and to using the same techniques to route claims to the appropriately skilled resources.
The second phase of the project integrated a real-time scoring solution with the existing first notice of loss process at Infinity Property & Casualty to achieve three primary objectives:
Shorten the claim cycle by handling more claims in the first notice of loss area without involving field adjusters.
Route claims with elements that might indicate fraudulent activity to the Special Investigations Unit at first notice of loss.
Route claims that require complex handling to a team with sufficient experience to handle that type of claims.
I was the lead Data Scientist for both phases and primarily responsible for the design and implementation of the data management as well as the business rules used in the evaluation of the fraud risk and the potential for subrogation for the individual claim.
I also worked closely with the business analysts at Infinity Property & Casualty to design the functional integration with the first notice of loss process as this included a number of new user interfaces for collecting additional data elements and identifying the points in the process that would trigger the real-time scoring.
As a result of the implementation, Infinity Property & Casualty doubled the accuracy of its fraud identification and achieved a 35% improvement in claims processing speed, contributing to a return on investment of 403%.
While the project at Infinity Property & Casualty is the project for which the most detailed account is available, I have worked or consulted on a number of similar projects in both the US and in Europe including an early implementation of the solution at Alka Forsikring in Denmark, which helped us further develop the software that formed the foundation of these projects.
Case studies and media coverage
Infinity Insurance staking the claims process on predictive analytics
Video featuring Director of Claims Analytics Mike Knabbb.
IBM helped Infinity Insurance transform the way agents handle and route claims, resulting in more satisfied customers. Predictive analytics is used to help Infinity address a number of challenges and opportunities. This transforms the way Infinity's agents handle and route claims, resulting in a lesser reliance on external adjusters, lower adjustment costs and - because claims are handled faster - more satisfied customers.
Video featuring Senior Vice President of National Claims Bill Dibble.
Smart is...combating insurance fraud with data. Infinity Property & Casualty, leveraged the IBM Big Data and Analytics platform to transform the way its agents handle and route claims. The solution resulted in a lesser reliance on external adjusters, lower adjustment costs and—because claims are handled faster—more satisfied customers.
IBM enables Infinity Property & Casualty Insurance to combat fraud
ROI case study
by Nucleus Research
Infinity Property & Casualty Corporation (IPACC) deployed SPSS to reduce its payments on fraudulent claims and improve its ability to collect payments from other insurance companies.
Kenneth A Jensen
A data science practitioner who will deliver results at every stage of the analytics lifecycle