For years, the commercial insurance team at Hummel Group relied on manual indexing and processing of our renewals. When a renewal policy came in, our quality control team would conduct screen-to-screen comparisons between the current policy and the renewal policy, looking for errors or unexplained differences.
This step is necessary to ensure our clients have accurate policies, as well as to protect our agency from E&O exposures, but the process was very time-consuming and difficult to scale. As the department grew – roughly doubling within the last five years – the team struggled to keep up with the volume of policies. Our leadership team at the agency started researching ways to make the process more efficient.
AI-backed insurance renewal processing
Our research led us to an AI-powered tool offered by a company called Exdion.
The tool is an artificial intelligence program that can compare current term policy forms to renewal policy forms with little manual input through an integration with our agency management system.
We assign a task to a bot user in our ImageRight software, asking it to compare the policies. It then extracts the PDF copies of policy forms and endorsements from the prior term, as well as the renewal policy, and feeds them into its policy comparison tool. The output is a spreadsheet that shows where coverage limits, schedules, forms, etc. did not match between the documents. The bot user pulls that spreadsheet back into ImageRight and assigns a task to the account manager, who looks over the results and takes necessary action from there.
This allows our team to easily address discrepancies without having to pore over every part of the policy manually.
The process of implementing an artificial intelligence tool
Implementing a new technology tool takes time, especially when it is a tailored solution. We worked with the Exdion team to tweak the tool to our agency’s specifications. There was a lengthy runway for that initial build.
We also had to train the bot on our agency’s naming conventions so it could know what documents and information to pull and what to ignore. Thankfully, our quality control team had established consistent data and naming conventions in our agency management system. This put us ahead of the curve and made training the bot much simpler.
Integrating the bot into our systems and training it took about three months start to finish. We then spent about a month testing it, then rolled it out to our team and trained them on how to use it.
Overall, we started the initiative with Exdion in June of 2023 and had the tool fully live by early 2024.
Getting the team on board
When we started introducing this new concept, we were very careful in how we communicated with our team. New technologies and processes can be intimidating, especially when a technology tool is significantly day-to-day workflows.
We knew that for our quality control team, their immediate thought might be, “I am losing my job. Technology is replacing me.” That was not our intent at all. That is not the way we operate. Rather, this created an opportunity for them to take on tasks and responsibilities that have a greater impact on our business and will allow them to continue to develop professionally within our organization. This team saw this, as well as the immediate workload relief, and quickly bought into the change.
Getting our account managers on board was more challenging. This team was used to the old way of doing things, and this change would bring a disruption to a comfortable workflow.
To secure buy-in, we had to find a few champions to support the new process early on and get involved in educating and training the rest of the team. It took a little while to get everyone on board, but once the team started seeing the proof of concept, they bought in.
The proof of concept
Renewals that would have sat in a backlog for weeks or months are now processed and back to the account manager within 72 hours.
By receiving processed renewal policies and quotes so quickly, our team can work much further ahead and be proactive rather than reactive with our clients.
Taking out the manual labor component of renewals has also allowed us to pivot our quality control team to tasks that go beyond renewal processing. This team can now support account managers with tasks such as quoting, endorsement and certificate processing, covering account manager duties over vacations, and more.
Our goal is to make sure our people have the tools and resources they need to be successful. This new process is another step towards that goal. The process can be scaled considerably, positioning us nicely for continued growth. Implementing an artificial intelligence tool to handle routine renewal tasks has freed up additional back-office resources and gotten us to a point where we are ready to keep this train rolling.
