Automating Claims Intake and Triage for Lakeshore Mutual Insurance's 300,000 Policyholders

Lakeshore Mutual Insurance serves more than 300,000 policyholders across the Midwest, but filing a claim in 2024 still meant picking up the phone, sending a fax, or mailing in paperwork — there was no digital self-service option at all. Every claim that came in was manually triaged and assigned to an adjuster by a claims coordinator working through a queue, which alone added days before anyone had actually reviewed the claim's substance.

Policyholders routinely waited over a week just to hear back after filing, which is a genuinely difficult experience when the claim in question is often about something stressful — storm damage to a home, a car accident, a burst pipe. Meanwhile, fraud detection relied entirely on adjuster intuition built from experience and after-the-fact audits; there was no systematic way to flag a suspicious claim pattern before the payout had already gone out.

Every status inquiry required a phone call, which meant the call center absorbed a steady stream of "what's happening with my claim" calls that a self-service system could have handled instantly — straining call center capacity and making policyholders wait even to ask a simple question.

We built a digital claims intake portal as the front door to everything else — policyholders can now file a claim online or via mobile in minutes, uploading photos and documentation directly rather than mailing them in. That structured digital intake is what made automated triage possible in the first place: the system now routes each claim to the right adjuster based on claim type, complexity, and current workload, work that used to require a human coordinator manually reading every submission.

Fraud detection was the piece Lakeshore's claims leadership was most cautious about, understandably — false positives that delay a legitimate claim create real harm. We trained a machine-learning model on Lakeshore's own historical claims data, looking for the patterns that had actually preceded confirmed fraud in the past rather than generic industry heuristics, and we ran it in parallel with the claims team's normal process for an extended validation period before it influenced any live triage decision.

Self-service status tracking closed the loop on the call-center problem: policyholders can now check exactly where their claim stands without calling in, which took a meaningful chunk of routine inquiry volume off the phone lines. We rolled the new system out by claim type rather than all at once — starting with simpler, lower-complexity claim categories where automated triage had the clearest track record, before extending to more complex claim types once the team had built confidence in the system's judgment.