Effi Mor, Co-Founder & CEO
The insurance industry is experiencing a paradigm shift. The advent of Artificial Intelligence (AI) and Machine Learning (ML) techniques is helping insurers to manage financial and actuarial risks efficiently. Israel-based creative tech startup, RemitRix, is upping the ante by harnessing the power of ML to help actuaries vastly improve their predictions.
Effi Mor, Co-founder of RemitRix and tech evangelist, describes the challenges follows, “actuaries hold a massive amount of data, which is not utilized properly due to the limitations of classical actuarial models. This inhibits their cash flow prediction accuracy, as they are limited in their ability to find small segments in the population with unique behaviours. This segmentation can improve not only the pricing process, which is relatively simple and straight forward, but can also help to better project the capital requirement.”
"Today, RemitRix offers state-of-the-art technology products and solutions for actuaries and insurers"
The Results Speaks for Themselves
Today, RemitRix offers state-of-the-art technology products and solutions for actuaries and insurers. RemitRix Agile and RemitRix Horizon are web-based financial and actuarial risk management platforms. RemitRix Agile includes classical-risk management tools such as ESG VaR, optimization tools and more. RemitRix Horizon will help actuaries to understand the contribution of different products to the capital requirements and focus on the business segmentation. It is, of course, designed to integrate with RemitRix Agile. Both products are easy to deploy and manage, are compliant
Additionally, RemitRix’s game-changing solution RemitRix ModuLearn helps insurers reduce their capital risk by better predicting their future costs and cash flows with the help of ML algorithms that utilize the company's data. Netta Shachar, ML researcher in RemitRix, explains one case study, “We have an XGBoost algorithm to predict the yearly claim rates for two health products: ambulatory insurance and surgery insurance. We were able to achieve a more accurate estimator for the claim rate in both products, along with a confidence interval and additional interesting insights into the modelling process and the data used.”
The predictions achieved using ‘XGBoost’ were closer to a true number of claims than those achieved by classical actuarial models for both products. For ambulatory, ML prediction overestimated by less than 1 percent, whereas the classical approach overestimated by almost 7.5 percent. For Surgery, ML underestimated by less than 1 percent while the classical approach overestimated by 7 percent. Along with these significant improvements in accuracy, using ML practices reveals additional insights to the health data which may prove valuable.