There Is More to CLAIMS Than Just “AI”

Aug 24, 2021

Theo Arends

Artificial intelligence (AI) has been the magic phrase for some years now, especially with fraud analytics during the claims process. And we, at FRISS, use different aspects of this; think text mining, predictive models, network analysis or image analysis. However, in general, the power of AI can be exaggerated. Blatant claims are often made about its abilities, with no evidence to back that up. Things like: “AI can easily be used anywhere without constraints”. And in our experience, this is not the case. Aside from AI, there are many other factors necessary for success, hence “more letters in CLAIMS than just AI”.


First, C is for core systems. The integration of fraud detection into core systems is essential for carriers’ success. Fraud detection should be seamlessly added to the daily workflow of adjusters, so what better way than through their core system? And, the results of fraud scoring should be clearly visible too. Actionable insights need to be offered to support adjusters’ decisions, and adjusters need the ability to deliver feedback on those insights. This system is essential for establishing straight-through processing and other automated processes. On top of that, the ability to transfer information between claims and SIU departments is critical. FRISS observed a need for this kind of technology and now offer it through our investigations solution. Integration into core systems is key from the perspective of efficiency, effectiveness, and overall fraud savings. Because of this, FRISS has invested a lot into partnerships with core systems vendors, building clear and intuitive integrations to create tools that facilitate daily work.


Next, L is for Lessons Learned. FRISS has already completed more than 200 implementations, and we’d be lying if we said everything went smoothly during these projects. Plenty of knowledge has been gained. Because of that, we’ve been able to take these lessons and use them as learnings to modify out methods for the future. For example, we now know that the transition from our product team to our customer success team is essential and needs extra attention from our side. In general, the more people involved, the greater the likelihood for real-time fraud detection to be adopted. Other knowledge we’ve gained is related to actionability (next steps, new fraud schemes) and the necessity for better visuals at screenings too. From the measures we’ve already taken, we can see the dramatic results in our customers’ success, and we continue to learn from their feedback. These lessons never stop and they’re truly what drive our ambition.


Next, M is for market knowledge. For an outsider, insurance may seem like a quite mundane industry, but truth be told, there’s many ever-changing factors that keep our lives interesting. What is a carrier’s risk appetite? What are their target groups and distribution channels? What does their portfolio look like? How is data compiled in all of these different countries and regions? All of these and more are very important questions, responsible for influencing how AI will be used in each specific company. So, business knowledge is an essential prerequisite for a SaaS company. Adjusters and investigators can’t be supported if we, at FRISS, do not fully understand them and their work. We need their insight to decide what the next steps are with the outcome of AI-fueled features.


Finally, S is for success! AI is a progressive mean for detecting fraud. Fraud models and image analytics are great features, but they should remain the means and not the end. Factors like integration, local modus operandi, domain knowledge,  and availability of data (sources) are just as important.  Claims still requires the expertise of adjusters and investigators to be truly effective. That paired with AI, is what leads to true success. As a success manager, I’m happy to discuss with you how you can position yourself to enable safe digital transformation and accelerate your business.