How we built a personalized health recommendation engine
Abstract: For a client in the health insurance domain, a health recommendation engine is crucial. We integrated this engine, after the design and development phase, on top of an existing health trajectory prediction application. Our goal was to ensure the client could offer recommendations on treatment and disease prevention by lowering their potential risk.
One of our health insurance clients is active in a competitive market where AI can be a true differentiator. This is an industry where limited customer interaction, and a lack of personalization, can lead to clients switching to a competitor with similar coverage options but more personalized services.
To this end, investing in AI solutions that support personalized services based on existing data is critical. Since our client was interested in their customers’ wellbeing, they wanted to implement a provider-customer interaction system that collects feedback—and therefore creates a more personalized health offering.
The client approached SPRYFOX to develop a recommendation system with a personalized health offering. Their desired outcome was, on the one hand, a better customer product experience, and on the other hand, a more accurate cost estimation for the provider.
Previous projects with the customer laid the foundation for this task. After a series of joint sessions, we developed a common idea for a recommendation system. Then, with the assistance of subject matter experts from the medical field, we defined the project so that the strong medical foundation translated into a technical implementation.
To create a customized experience, we used customer data, health history, and resulting health trajectory predictions. In addition, we integrated open data sources before linking them to the customer data. As a result, the technical solution integrates with existing applications, leveraging both customer data and previous projects.
The resulting application is embedded in the client’s IT landscape. A flexible API integrates tightly with existing products, while a corresponding web application showcases features and findings to enable data inspection on the fly.
Using the developed recommendation engine, our client can now provide personalized recommendations to the hundreds of thousands of customers who rely on them to increase their wellbeing.