Look Before You Leap: Implementing AI for Commercial Insurance

Artificial Intelligence - July 17 2019

According to a new report conducted by CompTIA, a mere 19 percent of companies say that they have expert knowledge around AI. This finding is based on a survey of 500 U.S. business and technology professionals. As David Weldon writes in Digital Insurance, the benefits of artificial intelligence are well known, but the adoption and use of AI technologies has been hampered by “confusion about how to best implement them.”

Given the lack of internal AI skills, it’s no surprise that the majority of companies opt to acquire AI capabilities rather than develop them in-house. According to Deloitte, 59 percent of companies get their AI through enterprise software vendors with specialized AI expertise. That being said, implementing an AI solution is a shared responsibility between the insurer and the insurtech partner providing the AI solution; both must work together and do their part to achieve a successful deployment. (There’s been quite a lot written lately about the importance of getting this insurer-insurtech partnership right.)

Before jumping into an AI initiative, commercial insurance brokers and carriers need to understand the various types of AI solutions available, the use cases for these solutions, and what is involved in implementing AI. While in many ways, deploying an artificial insurance solution is similar to rolling out any other insurance system, there are important differences that insurers need to understand and prepare for.

In an effort to help brokers and carriers get up to speed on the technical aspects of implementing AI, we’ve put together a new technology guide: Artificial Intelligence in Commercial Insurance.

Inside we look at different technologies including artificial intelligence, machine learning (ML) and natural language processing (NLP), and answer common questions brokers and carriers have about the workflows they enable, and what is required to implement them, such as:

  • What IT and business roles are needed on the broker or carrier side?
  • What are the phases and timelines for rolling out an AI solution like Chisel?
  • What type of data – and how much data – is needed to train the model?
  • How does the AI solution integrate with core insurance systems?
  • What are the technical KPIs for measuring the accuracy of an AI solution?

The aim is to cut through the confusion and provide a clear picture of what a real-world AI implementation looks like. We discuss where Chisel fits in the spectrum of AI solutions, as well as the broker and carrier workflows we digitize and automate. We also walk you through the various implementation phases – data preparation, training, audit and QA, production, and ongoing optimization – and discuss the roles and responsibilities involved at each step along the way.

Commercial insurance is complex, and the stakes are high when it comes to investing in AI. When brokers and carriers know what they’re getting into and what to expect, the AI implementation will invariably go more smoothly from start to finish. It’s critical for insurers and their insurtech partners to be on the same page when it comes to implementing AI and measuring outcomes.

Your insurtech partner possesses the expert knowledge around AI that you may be lacking internally – this is why you buy their solution and partner with them in the first place. Understanding what you can expect from your insurtech, and what they expect from you, will ensure that your deployment is as frictionless as possible, and help you realize a quick return on your AI investment.      

Get the Tech Guide: AI in Commercial Insurance

 

Browse different topics

Recent Posts