Issues & Answers Special Advertising Section
Head of Americas Technology Sales, Insurance Consulting and Technology
Willis Towers Watson
“Our Radar Base and Radar Live solutions deliver advanced analytic results.”
How are insurers benefiting from their investments in machine learning and data science?
With reduced compute costs and an automated model-development process, insurers can enable a more efficient data-science team, which results in a much more manageable investment. Using cutting-edge machine learning and artificial intelligence, the entire model-development process can be automated, or just certain components of it. This relieves the team from actually having to do the coding for any of the components that are automated. Willis Towers Watson is developing an automated machine learning tool that will do exactly this for the insurance industry. It’s a fully scalable
SaaS-based solution — giving insurers the ability to scale up and scale down the infrastructure they need, when they need it. Combined with automating the model-development process, this will enable insurers to optimize their investments and rewards from these initiatives.
What issues are insurers facing when it comes to deployment of rates, underwriting rules or general analytics?
The two issues insurers are facing quite frequently are speed to market and system constraints. One of the issues we see is that it takes weeks to implement a basic rate change and months to implement something that’s a bit more complex, like adding a new table. With this, we’ve seen cases where insurers will file rates, and then six months down the road they finally have the resources available to begin coding to implement the rate changes. Furthermore, there are cases where the pricing team develops the rates and, if they are more complex, in some cases the IT team cannot actually directly deploy those rates. They end up being dumbed down or shoehorned into the existing system. This is because of the constraints of the existing system.
How is Willis Towers Watson helping clients address these issues?
By leveraging our Radar Base and Radar Live solutions, we help clients avoid these and other bottlenecks in many ways. First, by eliminating the need for recoding, clients using our Radar Live solution can directly deploy the rates that the pricing team develops, so there’s no need for recoding and there’s no disconnect. Another way that we help clients is through simplifying the validation process. Instead of validating on a case-by-case basis, there is a systematic way that large numbers of online and offline results can be validated. This enables the rates and rules that are deployed to be deployed more correctly.