With the help of artificial intelligence, machine learning and the AWS Cloud, Policy Expert was able to get more insights towards its customers and more confidently drive decisions regarding the insurance quote features, thereby pushing innovation with an eye on enlarging their competitive edge.
Policy Expert is the trading name of QMetric Group Limited, an insurance company launched in 2010 which combines leading edge technology and innovative underwriting to offer competitively-priced products ranging from home insurance, motor insurance to travel insurance.
Policy Expert is a digital and technical stronghold and wanted to significantly expand their know-how related to setting up a Machine Learning environment, best practices and state-of-the-art modelling techniques. These components are essential to accommodate the level of AI needed to generate customer related insights and radically innovate a digital insurance quote’s lifecycle on multiple fronts, including real-time decision making on the insurance policies they offer.
Chief Pricing Officer
Christine Minetou (Policy Expert)
By leveraging AWS components such as AWS Glue for (timely) data processing, AWS SageMaker for model training & hosting and AWS Fargate to host our applications, we co-developed an MLOps solution for a first use-case. With Thompson Sampling at the basis, we optimised the exploration and exploitation balance that is inherent to sequential real-time quote generation.
Working with Policy Expert was rewarding. Combining our ML and AWS capabilities with their technological know-how and business understanding, we were able to go to production in the shortest cycle possible and make the right decisions along the way. - Florentijn Degroote (Senior Machine Learning Engineer at ML6)
The ML6 machine learning experts helped the internal team create a clear project roadmap to validate assumptions, generate data driven insights and drive decision making. After an advisory track, ML6 was able to create a good understanding of the challenge and could map out an initial technical proposal, highlighting the important caveats. After the workshops, ML6 helped structure the data and project by explaining the concepts in simple terms, with concrete examples, and decomposing the project into different steps. Based on the advice from ML6, Policy Expert got to work hands-on with MLOps and ML-based AWS components, allowing them to significantly speed up the innovation cycle.