Claims

Problem

Claims were difficult to process efficiently, not because of missing tools, but because of how work moved through the organization.

Much of the claims intake and review process was handled manually. Information moved through calls, emails, and multiple handoffs, creating bottlenecks and delays — even for straightforward claims.

In some cases, lower-risk claims under established thresholds could be paid out quickly, but were still slowed by the same review paths as more complex cases. This resulted in unnecessary hold-ups, inconsistent timelines, and frustration for both claim handlers and stakeholders.

Research Insights

Research focused on understanding how claims actually moved through the organization rather than how they were documented.

• Low-risk claims were often over-processed despite clear payout thresholds.
• Decision authority was unclear, leading to unnecessary escalations.
• Manual communication channels created visibility gaps in claim status.
• Similar claims were handled inconsistently, introducing delays and unpredictability.

These issues were not the result of individual mistakes, but of a system that treated all claims the same regardless of risk or complexity.

Design Challenge

The challenge was not to digitize the process, but to align it with how claims should be handled based on risk and complexity.

• Support faster paths for low-risk claims without compromising compliance
• Clarify ownership and decision points across roles
• Reduce unnecessary handoffs and follow-ups
• Work within existing policy, system, and regulatory constraints

My Role

I led the product design effort end-to-end, including research synthesis, workflow mapping across roles, information architecture, interaction design, and UI design.

I partnered closely with product, engineering, and subject-matter experts to validate assumptions and ensure design decisions aligned with operational realities.

Process

I began by mapping the current-state claims workflow across roles, decision points, and communication channels. This surfaced where claims were treated uniformly despite having very different risk profiles.

I then designed a future-state flow that introduced clear decision points early in the process. Based on claim characteristics and thresholds, the workflow could follow different paths while remaining part of a single, cohesive system.

The redesigned flow clarified where decisions should be made, who owned each step, and which claims required deeper review versus minimal friction.

Risk-based workflow showing how early decisions route claims through faster or more detailed paths based on complexity and thresholds.

Key Screens

The interface design focused on the primary claim intake task. The goal was not to automate judgment, but to support it by structuring information in a way that aligned with downstream decisions.

The core task screen demonstrates how relevant information is grouped by decision context, requirements are made explicit at the moment they matter, and different claim paths are supported without unnecessary steps.

The Details step captures classification and loss information early, setting up downstream steps and reducing rework later in the process.

Solution

The final solution aligned the claims experience with how work actually needed to happen.

By introducing clear decision points and conditional paths, the workflow supported faster handling of low-risk claims while preserving appropriate review for more complex cases. Manual communication was reduced by making ownership, requirements, and next steps explicit within the system.

Outcome

The redesigned workflow reduced unnecessary bottlenecks and improved consistency in how claims were handled across the organization.

Claims that could be resolved quickly were no longer slowed by processes designed for higher-risk cases, while more complex claims retained the oversight they required.

With more time, I would focus on adding instrumentation to measure throughput, decision time, and handoff frequency, and continue refining edge cases uncovered through ongoing use.