Work

About

Designing for Scale: Exception-Based Automation for HR Workflows

Motus is the industry leader in vehicle reimbursement, offering solutions that extend beyond mileage to personal asset reimbursement - including mobile devices, internet, and home office setups. During the pandemic, we launched a new product called BYO (Bring Your Own) to meet the evolving needs of distributed workforces.

A unique project where I wore two hats. As APM I drove the product strategy, and led our new contract designer to translate that strategy into an admin-facing interface that reduced manual claim reviews and surfaced only the exceptions that truly mattered.

① PROBLEM STATEMENT & GOALS

① PROBLEM STATEMENT & GOALS

① PROBLEM STATEMENT & GOALS

Manual reviews at scale create noise, and noise creates errors.

Companies using Motus to manage reimbursements for cell phones, home office equipment, and internet costs were putting their admins through a punishing monthly ritual: manually reviewing every single claim, even when the vast majority followed predictable, already-approved patterns. The signal-to-noise ratio was terrible. The real issues (new addresses, program changes, leave dates) were buried.

② DESIGN PROCESS

② DESIGN PROCESS

Although we had a contract UI designer to help execute on visual layouts, I led the user discovery and solution definition, ensuring that our designs were rooted in admin needs and business goals. Our process included:

  • Discovery: 1:1 interviews with admins to highlight pain points and identify patterns in exceptions

  • Ideation: Defined automation logic and workflows in partnership with Engineering

  • Prototyping: Collaborated with the designer to translate workflows into interactive prototypes

  • Validation: Ran usability tests with real admins to confirm the solution was intuitive and effective

③ REQUIREMENTS TO SOLUTIONS

③ REQUIREMENTS TO SOLUTIONS

My Role

I led this project as Product Manager, partnering with a brand new contract designer who was unfamiliar with the Motus product and users. I owned:

  • Gathering client feedback and validating concepts

  • Defining the automation rules and exception logic

  • Prioritizing features and writing user stories

  • Coordinating with Engineering through delivery and QA

The screens here are my own redesign, applying our current design system to those same core product decisions. I'm including this case study to show I can think beyond mobile consumer, specifically around dense data tables, permission systems, and UI where clarity matters more than delight.

Key Capabilities

01. Dashboard-Level Alerts: Admins now saw a summary banner on the dashboard notifying them of pending exceptions. This reduced the need to click into each report and helped prioritize their attention.

The dashboard notification gives admins immediate signal at the start of the pay period. Rather than navigating into the report blindly, they land with full context: how many total claims exist, how many were auto-approved, and exactly what exception types demand action. The breakdown in the alert prevents surprise.

02. Customizable Approval Settings: We introduced a flexible settings page allowing each company to select which exception types they wanted flagged. This ensured the system could adapt to varying internal policies and workflows.

The settings surface gives enterprise admins meaningful control without overwhelming them. Each exception type is a binary toggle with a plain-language description. Companies that don't care about rate change flags can turn them off. The system adapts to their risk tolerance, not the other way around.

03. Inline Reimbursement Report Enhancements: To reduce cognitive load, flagged claims were surfaced directly within the reimbursement report, keeping admins in context and minimizing navigation friction. Rather than reviewing all claims, the system automatically flagged only those with specific exceptions. We encoded five exception types that aligned with client expectations and compliance needs, ensuring only high-risk claims were reviewed.

The enhanced report keeps admins in their primary workflow. Flagged claims float to the top when the exceptions filter is active. Inline approve/deny actions eliminate the need to navigate to a separate review screen. Auto-approved rows are visible but visually de-emphasized, preserving context without creating noise.

The dashboard notification gives admins immediate signal at the start of the pay period. Rather than navigating into the report blindly, they land with full context: how many total claims exist, how many were auto-approved, and exactly what exception types demand action. The breakdown in the alert prevents surprise.

Key Capabilities

01. Dashboard-Level Alerts: Admins now saw a summary banner on the dashboard notifying them of pending exceptions. This reduced the need to click into each report and helped prioritize their attention.

The settings surface gives enterprise admins meaningful control without overwhelming them. Each exception type is a binary toggle with a plain-language description. Companies that don't care about rate change flags can turn them off. The system adapts to their risk tolerance, not the other way around.

02. Customizable Approval Settings: We introduced a flexible settings page allowing each company to select which exception types they wanted flagged. This ensured the system could adapt to varying internal policies and workflows.

03. Inline Reimbursement Report Enhancements: To reduce cognitive load, flagged claims were surfaced directly within the reimbursement report, keeping admins in context and minimizing navigation friction. Rather than reviewing all claims, the system automatically flagged only those with specific exceptions. We encoded five exception types that aligned with client expectations and compliance needs, ensuring only high-risk claims were reviewed.

The enhanced report keeps admins in their primary workflow. Flagged claims float to the top when the exceptions filter is active. Inline approve/deny actions eliminate the need to navigate to a separate review screen. Auto-approved rows are visible but visually de-emphasized, preserving context without creating noise.

My Role

I led this project as Product Manager, partnering with a brand new contract designer who was unfamiliar with the Motus product and users. I owned:

  • Gathering client feedback and validating concepts

  • Defining the automation rules and exception logic

  • Prioritizing features and writing user stories

  • Coordinating with Engineering through delivery and QA

The screens here are my own redesign, applying our current design system to those same core product decisions. I'm including this case study to show I can think beyond mobile consumer, specifically around dense data tables, permission systems, and UI where clarity matters more than delight.

Key Capabilities

01. Dashboard-Level Alerts: Admins now saw a summary banner on the dashboard notifying them of pending exceptions. This reduced the need to click into each report and helped prioritize their attention.

The dashboard notification gives admins immediate signal at the start of the pay period. Rather than navigating into the report blindly, they land with full context: how many total claims exist, how many were auto-approved, and exactly what exception types demand action. The breakdown in the alert prevents surprise.

02. Customizable Approval Settings: We introduced a flexible settings page allowing each company to select which exception types they wanted flagged. This ensured the system could adapt to varying internal policies and workflows.

The settings surface gives enterprise admins meaningful control without overwhelming them. They can select which reasons are most relevant to their company policies. Companies that don't care about rate change flags can turn them off. The system adapts to their risk tolerance, not the other way around.

03. Inline Reimbursement Report Enhancements: To reduce cognitive load, flagged claims were surfaced directly within the reimbursement report, keeping admins in context and minimizing navigation friction. Rather than reviewing all claims, the system automatically flagged only those with specific exceptions. We encoded five exception types that aligned with client expectations and compliance needs, ensuring only high-risk claims were reviewed.

The enhanced report keeps admins in their primary workflow. Flagged claims can be viewed in a dedicated tab, and filters give the admin precision when approving their monthly report.

My Role

As the Product Manager, I led discovery, strategy, and delivery—partnering with Engineering, Customer Success, and a contract UI designer to bring this feature to life. I owned:

  • Defining the automation rules and exception logic

  • Gathering client feedback and validating concepts

  • Prioritizing features and writing user stories

  • Coordinating with Engineering through delivery and QA

Later, during my time as a Product Designer, I helped implement our internal design system and updated the visual layer of BYO features—bringing cohesion and consistency to the experience.


④ KEY RESULTS

④ KEY RESULTS

Smart Approvals had a measurable impact on both efficiency and user satisfaction:

  • Cut down manual reviews by allowing most claims to be auto-approved

  • Improved admin experience with clear exception alerts and contextual design

  • Increased trust and adoption of the BYO product among larger clients with complex workflows

Admins reported feeling more in control and less burdened by repetitive, low-value tasks.

⑤ CONCLUSION

⑤ CONCLUSION

This project solidified my belief in automation as a driver of business value, especially when paired with flexibility and thoughtful UX. As both a PM and later a Designer, I learned how critical it is to deeply understand user pain points, but also to champion scalable, configurable solutions that work across different customer needs in order to significantly enhance satisfaction and efficiency in enterprise workflows.

SAY ANNYEONG!