In the era of AI, insurance agencies have moved past debating the merits of the technology. The next big question is “How are we using this technology to save us time and make room for more valuable work?”
During a recent ACN Epic Power User Ask Me Anything session, agency leaders shared practical lessons on how they are using AI to reduce administrative work, improve efficiency and support staff without sacrificing human oversight.
AI Is Already Working Inside the Agency
Jordan Smith, CIO of The Hartwell Corporation, shared that his organization uses a variety of Applied AI capabilities and third-party solutions across multiple departments and workflows.
Those applications include Bridge for email summarization; Coverflow for proposals, policy comparisons and gap analysis; Quandary for automated personal lines remarketing; Ascend for accounting workflows, such as direct bill reconciliation and payables; KAI for benefits administration tasks, including employee enrollment changes and policy renewals; and Microsoft Copilot for general productivity use.
These tools are being engaged to solve specific operational challenges. “We started by looking at different workflows that AI could help us get the most return right off the bat,” Smith said.
Start With Your Foundation
One of the strongest lessons shared during the discussion was that agencies should prepare their data and processes before implementing AI.
“When you want to go down that path, you look at setting up your own house first. We made sure we had our processes documented and our data fairly clean,” Smith said.
To effectively implement AI, he advised agencies to review and improve their operational readiness, which will become the foundation for implementation success.
Let Employees Identify the Best Opportunities
Smith indicated that one of the best ways to identify where AI can provide the most value is to seek advice from the insurance agent on the ground. These are the employees doing the day-to-day work.
According to Smith, agencies should survey employees in each department to determine their pain points. He adds that the survey can ask two simple questions:
- Which tasks take the longest?
- How much time does each task take?
The exercise often reveals that management's assumptions about inefficiency don’t always match the employee experience. By identifying the tasks that consume the most time, agencies can prioritize AI initiatives that deliver measurable productivity gains.
For Smith’s agency, proposal generation was considered a pain point for many employees. In commercial lines, proposal preparation often required substantial cleanup of information generated from Applied Epic. Depending on the complexity of the account, the process could take anywhere from 10 minutes to several hours.
Those findings led the agency to implement Coverflow to accelerate proposal creation and reduce manual effort. By starting with the bottlenecks identified by agents, the agency was able to identify the best workflow for automation and improve productivity and efficiency.
The Human-in-the-Loop Still Matters
As AI capabilities continue to improve, participants repeatedly emphasized that human oversight remains critical.
According to Smith, one of the biggest misconceptions about AI is the expectation that it should perform perfectly from day one.
“The AI tools available today are not meant to be 100% right,” he said. “We might get you 90% there, and then you're the person that goes through and makes sure it's complete.”
That philosophy was echoed throughout the discussion. Whether agencies are using AI for policy checking, proposal generation, coverage comparisons or remarketing and benefits administration, licensed insurance professionals remain responsible for reviewing outputs and validating results.
“The buck stops with the licensed agent,” noted one participant.
As AI capabilities expand, governance is becoming just as important as innovation.
Smith encouraged agencies to establish acceptable-use policies, security guidelines, data governance standards and compliance controls. He also highlighted tools, such as Microsoft Purview, which can help organizations prevent sensitive information from being inappropriately shared with AI systems.
Change Management Is Often the Hardest Part
While technology implementation requires effort, participant experiences suggested that user adoption is frequently the larger challenge.
During the AI rollout, according to Smith, some employees were eager to experiment with the new technology, while others remained hesitant to change their established workflows. During the early implementation of AI-generated proposals, staff frequently pointed out imperfections in the outputs. Smith said that the response was not to abandon the technology but to set realistic expectations.
Over time, the agency invested in template customization, individual training sessions, employee feedback and workflow refinement. As users became more familiar with the technology, adoption improved significantly. In fact, Smith noted that subsequent AI implementations became much easier because employees were already comfortable working with AI-assisted processes.
His advice to agencies just beginning the journey is simple: “You have to experiment and use AI. Just get used to it.”
Integration Drives Adoption
When it comes to workflow integration, AI solutions embedded in Applied Epic “are a lot easier to integrate with your staff,” Smith said.
Rather than requiring employees to log into separate systems, the most successful tools tend to work behind the scenes, surfacing information where users already spend their time.
One example involved automated personal lines remarketing.
When a renewal download enters Epic, the AI solution automatically retrieves the policy information, submits it to multiple carriers, returns comparison results and places them directly into an Epic Activity. By the time a team member reviews the account, much of the preliminary work has been done.
“They just go into Epic, see that it's already been re-quoted, and they can have that conversation with the client,” Smith said.
Looking Ahead: Agentic Workflows and Submission Management
Participants expressed significant interest in several emerging capabilities from Applied. One of the most discussed was Submissions Manager.
According to Smith, the solution centralizes submission activity into a single view while leveraging AI to extract data from emails, automatically attach communications, update submission statuses and track carrier responses. For agencies managing large volumes of remarketing and submissions, the potential visibility gains could be substantial.
Another area generating excitement is submissionless renewals. The vision is a future where agency data can proactively generate quote options and place them directly into workflows before the renewal process begins.
“If more carriers are added to that list, you could imagine 10 or 15 quotes just waiting for you,” Smith said.
While many of these capabilities are still evolving, they illustrate how quickly AI is moving beyond chat-style interactions into end-to-end workflow execution.
The message from experienced users was consistent: Successful AI adoption begins long before an agency deploys its first tool. Agencies with a strong foundation of clean data, defined workflows, realistic expectations, employee engagement and human oversight can take full advantage of AI to improve workflows and reduce administrative work, enabling agents to focus on what matters — their customers.
Explore the ACN Learning Center for recordings, training resources and upcoming live sessions designed to help agencies get more value from their data and workflows.