January 15, 2020 by David Gambrill, Editor in Chief
To cope with thinning talent in a greying claims workforce, insurance organizations are starting to digitize their claims triage and other workflows.
It’s called “intelligent process automation (IPA),” a combination of automation, workflow processes, and artificial intelligence (AI) or machine learning.
Basically, IPA learns from a company’s rules and claims patterns to direct certain claims to adjusters who have the most experience, skills or talent in a particular area. “In any organization, there will be adjusters with different skills sets and experience that can handle certain types of claim files more efficiently, depending upon what the injury or claim profile may be,” explains Mathew Raymond, founder of Expeflow, a cloud-based, centrally-administered claims management platform for medical and healthcare claims.
To explain how it works, Raymond cites an example from Health Claims for Auto Insurance (HCAI), an electronic system for transmitting auto insurance claim forms between insurers and health care facilities in Ontario. Third-party health care providers in the province are required by law to submit their treatment plans for insurance claims to HCAI.
“We have all of these injury codes coming in from HCAI that have been entered into the insurer’s file,” as Raymond explains. “Now you can learn which adjuster is handling which types of files, depending on the type of the insurance policy. IPA can help insurers direct the right file to the right adjuster. For example, maybe you have a claimant come in and an emotional impairment is an aspect of the file. That can now be handled by someone who has more experience with those types of claims.”
The AI component takes IPA far beyond simple process automation, according to Expeflow CEO Terry Stepien. A standard example of simple process automation is when an insurance organization takes a fairly straightforward form and completely automates the process for collecting the data. No one in the company touches the form during the process; the information gets processed very quickly.
Step 2 of IPA is when an organization maps out its workflows, identifying who does what. Here, the knowledge transfer from senior to junior adjusters is essentially digitized. Stepien describes Step 2 as follows: “Now you can have adjusters with two to four years of experience doing things that senior adjusters who have 15-17 years of experience do. That’s because you’ve put the sequence of steps in that the [senior adjusters] may have had in their head, you’ve captured some of that knowledge and institutionalized it in the insurer, and now you can have more consistency across the force.”
AI and machine learning come into play during Step 3 of IPA.
“Once you have the [business process] information in place, the system can start suggesting things to you that you may not have thought of,” says Stepien. “The system can start tracking best practices. It may say, ‘In other situations like this, you’ve done this with a person,’ or ‘This is the kind of question that you [as an adjuster] might ask.’ So, as an adjuster, you can see what’s coming next and start to get ahead of it.
“The real benefit of AI is looking at all of your knowledge and, based on the rules you give it, the system can say, ‘Here is the information that may be relevant to what you are looking at today,’ as opposed to the adjuster trying to find the little missing nugget in a vast sea of information.”