Cut claims cycle time from weeks to hours. Underwrite smarter. Retain more policyholders.

A regional retail chain's. commercial insurer, a mid-sized property and casualty carrier serving small and mid-sized business clients, approached Vibba with a claims cycle time problem that was showing up directly in customer satisfaction scores and, increasingly, in renewal rates. Straightforward property damage claims, the kind that should have been resolvable within days, were routinely taking three to four weeks from report to payout, a delay policyholders consistently flagged in post-claim surveys as their primary source of dissatisfaction.

The problem in detail. Claims were reported by phone or through a basic web form, then manually entered into the carrier's claims management system by an intake team before being assigned to an adjuster, a process that alone could take one to two days before any actual investigation began. Adjusters then scheduled in-person property visits to assess damage, adding further delay depending on adjuster availability and geographic coverage, particularly in the carrier's more rural service territories. Underwriting for new commercial policies relied on a standard set of application variables reviewed manually, a process that took several days per application and, leadership suspected but couldn't confirm without better data, was both approving some higher-risk accounts and declining some genuinely low-risk applicants who didn't fit the traditional variable profile.

Implementation. Vibba deployed the AI FNOL system first, integrated with the carrier's existing claims management platform, automatically structuring claim data on report and routing straightforward property damage claims into a computer vision-assisted assessment workflow rather than the standard manual queue. Policyholders reporting a claim were guided through submitting photos of the damage directly through the carrier's existing mobile app, with the computer vision model assessing severity and estimated repair cost automatically. In parallel, Vibba deployed the AI-assisted underwriting model, trained on the carrier's historical underwriting and loss data, to evaluate new commercial applications against a broader variable set, with every recommendation still reviewed by an underwriter before a final decision.

Deployment and staff training. Adjusters received training on reviewing AI-assessed damage estimates and the criteria for escalating a claim out of the automated workflow into standard manual investigation when appropriate. Underwriters received training on interpreting the AI model's risk assessment output alongside their existing underwriting judgment, with the tool positioned explicitly as a decision-support input rather than an automatic approval or decline mechanism.

Results. Straightforward property claims processed through the new computer vision-assisted workflow moved from report to payout in a matter of days rather than three to four weeks, consistent with the 60 to 80% FNOL automation rates reported industry-wide within six months of comparable deployments. Post-claim customer satisfaction scores for claims processed through the automated workflow improved measurably compared to the prior manual process. On the underwriting side, the carrier began tracking approval accuracy against the broader risk variable set the AI model evaluated, and early results supported leadership's suspicion that the previous manual process had been both approving some avoidable risk and declining some genuinely low-risk applicants.

Long-term improvements. Based on these results, the carrier expanded the computer vision assessment workflow to auto claims as well, and has since used adjuster capacity freed by claims automation to focus more attention on complex, high-value claims that genuinely require in-depth investigation. Underwriting leadership now reviews AI-flagged variable patterns as a standing input into periodic underwriting guideline reviews, a level of data-driven insight the manual process had not previously supported.