Designing Trust and Adoption in AI-Assisted Claims Processing
The Project: The Intelligent Digitization and Validation System (IDVS) is an AI-powered platform that helps claims associates process insurance claims by extracting, validating, and structuring information from First Notice of Loss documents submitted through emails, scanned forms, and handwritten submissions. While the underlying AI significantly reduced manual data entry, associates struggled to trust and effectively work with the system's recommendations. The experience created frustration, duplicate work, and low adoption, limiting the business value of the platform. The goal was not simply to improve usability, it was to improve how humans and AI worked together within a complex claims workflow.
The Challenge: The claims experience was fragmented, manual, and inconsistent across roles, leading to delays, miscommunication, and user frustration. There was no shared understanding of the end-to-end journey or the critical “moments that matter,” making it difficult for the organization to prioritize improvements or modernize the experience effectively.
My Role: As User Researcher and Product Designer, I led discovery efforts to understand associate pain points, conducted heuristic evaluations of the AI-assisted workflows, and developed design concepts that improved usability, transparency, and confidence in AI-generated outputs. I also partnered closely with engineering teams to assess implementation effort and prioritize improvements for future releases.
Our Approach:
Conducted user interviews with claims associates to understand friction points within AI-supported claims processing workflows.
Performed a heuristic evaluation to identify usability issues impacting trust, efficiency, and adoption.
Created design concepts and workflow recommendations that simplified interactions between users and AI-generated claim data.
Collaborated with engineering teams to assess level of effort and prioritize enhancements based on business value and implementation complexity.
Deliverables
23 prioritized UX and visual design recommendations to improve the AI-assisted claims experience.
Helped product and engineering teams prioritize enhancements using implementation effort and business impact criteria.
Improved the organization's understanding of the relationship between AI explainability, trust, and adoption.
