I'm Ben! đź‘‹
I was brought in to the insurance solutions team for this project specifically for my skills of bringing complex problems and regulation requirements into “comprehensive” user interfaces that address issues through a friendly and delightful experience.
We partnered with an insurer that was looking at AI to be able to uplift their commercial underwriting teams because they we’re having difficulties processing all their underwriting requests in time due to a lack of resources resulting in a loss of business opportunities.
Through our conversations with the stakeholders, we quickly integrated a concept that was important for our team which was to keep the human at the center of the decision-making when AI would be involved. This was especially true in the context of insurance coverage, where a simple mistake could mean very important losses for the insurance company.
While doing some shadowing of the underwriters’ work processes, we discovered multiple ways in which we could improve the underwriting experience as a whole, even without the use of AI but simply being bringing the user’s workflow to another level. Their current softwares were not internal tools and were provided by monolithic enterprises that couldn’t accommodate the specific business needs of our client. By doing so we could then give underwriters more tools, information and AI insights to have a comprehensive view to take better decisions.
The mean pain point of the executive level stakeholders was the throughput of their underwriting department. They wanted to preserve a similar workflow and not compromise on the quality of decisions and were very confident in their team’s capabilities and effectiveness, but there simply weren’t enough resources the continually source the team with that could ramp up efficiently enough.
The solution needed to be scalable to gradually increase the amount of applications entirely or partially through AI and be easy to tune over time to reflect a change in risk appetite or marketing perspectives.
We opted for a standalone platform dedicated for underwriters that would be connected to their current insurance application data streams and would act as their marketplace for the inbound applications from all insurance brokers.
Users would have an overview of all applications that have come through and at which step they were halted at depending on the information that was either accepted by the AI model, rejected or need an underwriter review.
Focusing underwriters’ expertise where AI fails to process applications automatically.
AI is used to sift through similar applications as references to be able to compare against and adjust the coverage in a snap, a task which could take hours for a human.
Each step of the underwriting process can be modified granting management more control over the automatic treatment of different insurance applications as needs change.
