A decade overdue

Targeted the three highest-cost UX issues

The Health & Benefits mobile app hadn't been rebuilt since 2018. When the company budgeted for a modernization, I was selected as lead designer based on my work on the registration redesign and H&B rebrand.

I scoped the project around the three highest-cost UX problems:

1. Login sent users to Safari, which made the app feel broken on first touch.
2. Home was visually crowded with 15+ account types competing for attention — users couldn't find what mattered.
3. Claims took 15 taps, which pushed users to call support instead of completing in-app.

The app sat at 2.7 stars, and Aetna had made modernization a requirement. These were the highest-leverage problems to solve first.

Moved login back in-app

I removed the browser redirect first, because trust starts at login.

I redesigned the flow around a native Ping Identity experience: username/password, OTP, and biometrics inside the app. I designed the ideal state before the Ping POC was complete, then aligned it with technical constraints as engineering finalized implementation.

The result was a login that felt native, predictable, and secure.

Old home

The old home screen forced users to read before they could act. Key balances, alerts, and actions all competed at the same visual weight.

New home

The dashboard had to display data for 15+ account types — HSAs, FSAs, COBRA, commuter benefits, emergency savings — each with different primary data points. I didn't know what to prioritize because that knowledge lived with different product managers across the company.

So I ran a cross-functional workshop. I brought PMs from each account type into a structured Miro session, had them bring FullStory behavioral data, and worked through four categories: data requirements, design questions, technical constraints, and open forum. After the workshop, I sent a follow-up survey and synthesized everything into a summary for leadership alignment.

Two insights shaped the final design. First, debit card actions were the #1 support call driver — so I surfaced card status and actions prominently on the dashboard to deflect calls. Second, multi-account holders were actually rare. That meant I could optimize for the common case instead of over-designing for edge cases that looked important internally but barely existed in the real user base.

From there, I built a core dashboard card component with variants per account type, each showing the primary data points and urgent actions the workshop identified. Key balances come first, supporting data is grouped, and nudges make next steps obvious.

Old account details

The old account details view had the right information but weak structure. Basic tasks took longer than they should because users had to parse dense content.

New account details

I reorganized account details by user intent: what matters now, what users can act on, and what's secondary. That made comparison and decision-making faster on mobile.

Old transaction details

Old transaction details had low signal-to-noise. Users could find answers, but status and next steps were unclear.

New transaction details

I tightened hierarchy, labels, and status treatment so users can understand transaction state in one pass without losing depth.

Old vs new, side by side

This comparison shows the core shift: from dense screens that require interpretation to screens that support quick decisions.

In December 2025, I ran 8 moderated Lookback sessions with a Figma prototype. Dashboard scored 4.38/5. Consistent feedback: clean, intuitive, easy to navigate.

The sessions gave me clear direction on three things:

1. "Pay a provider" was unclear — I renamed the action to make both intents explicit.
2. "Investment balance" lacked context — I added clearer framing and entry points.
3. Users expected transaction history in different places — I improved the information architecture cues.

Cut claims from 15 taps to 3

Claims was the highest-intent task and the highest-friction flow.

I redesigned it as photo-first, with a shorter and clearer path to completion. I also unified claims and card activity into one recent-activity feed with readable status chips.

The goal wasn't just fewer taps. It was higher in-app completion and lower support volume.

Delivered an Aetna-ready prototype

For the February 2026 Aetna deadline, I delivered an end-to-end prototype of login, home, claims, and debit card management.

It wasn't pitched as finished UI. It showed that product direction, interaction quality, and execution were ready for modernization at scale.

Results that mattered

1. Login moved from a Safari redirect to a native in-app flow — the single biggest trust fix.
2. Claims went from 15 taps to 3, designed to shift users from calling support to completing in-app.
3. Dashboard scored 4.38/5 across 8 moderated Lookback sessions.
4. Delivered a full prototype that met the February 2026 Aetna modernization deadline.

What I'd ship next

1. OCR auto-fill for claims. I already prototyped this — a working interactive prototype built with Claude Code where users scan a receipt and OCR auto-fills the claim details. It's scoped into the product roadmap.
2. Passkey login as soon as backend support is ready.
3. Connected claims for faster adjudication and clearer post-submission feedback.

The redesign got the foundation right. These would take it further.