Insight
Big In Japan: AI in the Insurance Industry
March 15, 2024
One thing we hear a lot of are questions around use cases for AI in specific industries. Recently, NextAccess sat down with Wil Bachman of staffing firm Umbrex to discuss a consulting engagement in the automotive sector that used AI tools to streamline and improve the auto accident claims process.
The client, a global insurance conglomerate, was under pressure to reduce costs in its auto business in Japan. Given our role as leading the data science aspect of this project, NextAccess had visibility across all the different functional areas within the auto division and as such was able to identify claims as a high-expense area that could benefit from the implementation of advanced analytics and AI technology. As is often said within the insurance industry, claims are where the big checks are written, and such was the case for this particular client. Anything that could be done to improve performance within this area would be an obvious big win for them.
What was obvious here was that while the claims processes were highly structured and handled with great attention to detail, they weren’t designed around an efficient utilization of resources. This meant that claims were processed in the same manner regardless of the size of the individual claim, with similar levels of work put towards them whether they were valued at $500 or $50K. By applying analytical rigor to the claims estimation process, not only would this lead to a savings on processing effort but in response time to customers as well.
After identifying the issue, the next step involved taking a closer look at information flows at different points in the process, and specifically at the information collected at the time of the accident to see what insights could be gleaned. NextAccess then used this info to build predictive ML models using Python that allowed us to analyze historic data to determine the likely ultimate dollar value of the claims. These models enabled setting thresholds to separate out smaller claims suitable for automated processing vs larger and more complex ones that required greater oversight.
The results from this analysis and implementation had an immediate impact on the firm’s bottom line. With automation handling about 25% of the total number of claims, this reduced workload for the claims department and decreased expenses related to staff hours and administrative tasks. The accelerated resolution times for eligible claims also increased customer satisfaction, as people no longer had to wait an uncertain and lengthy amount of time to receive their payments. Through ongoing monitoring, the client was able to watch for any potential drifts in model performance and adjust accordingly. In the end, by embracing the use of AI technology in this fashion, the insurance conglomerate was able to realize cost savings in a manner that was not disruptive to the overall business but at the same time put them ahead of the curve with regard to their competitors.
