Soleil: designing for a job that happens between glances.

The most important decisions in this project came from watching people, not screens. I observed representatives using Soleil in the actual lounge environment, and the details reshaped everything. Unlike a retail cashier, reps configured their workstations to their own ergonomics—the variation between them was extreme. And the Admirals Club customer expects a higher tier of service, which hinges on something small and human: the rep making eye contact and smiling. That meant the real interaction wasn't rep-and-screen; it was rep glancing between screen and guest, over and over.

Many reps wore reading glasses—putting them on to read the interface, taking them off to make eye contact with the guest. The interface was fighting the service. So I designed the Tile System: large, icon-driven tiles that let a rep navigate to a task without reaching for their glasses, that worked for a tap-only flow on the tablets and phones the club wanted to adopt, and that—once learned—let reps find an action by symbol faster than by reading. The logic was everyday: you spot a restroom by the icon long before you read the word.

Interviews distilled into two personas that captured a real tension: Evette, the veteran who wanted the tool to get faster but didn't want to relearn patterns, and Emma, the newcomer open to entirely new hardware and flows. Designing for both meant improving performance without forcing veterans through a disorienting rebuild—which shaped a deliberately incremental migration strategy. Auditing Soleil surfaced structural problems, not cosmetic ones: language that spoke AA's internal procedures rather than the reps' actual tasks, and a navigation split between "Guest Registration" and "Miscellaneous Guest Registration" when there's no such thing as a non-guest registration. I set the priority order explicitly: put good service and the customer's experience first, then fit policy into that—and, since 90% of use was scanning or keying in guest data, built a dedicated registration field right into the header, cutting several steps out of the core task.

Line Buster: registering guests fast, before the line builds.

Soleil Line Buster was the mobile companion to the broader redesign—a stripped-down tool built for one job: registering a waiting guest as quickly as possible, so the line at the door never becomes the guest's first impression. Interviews made the real problem clear: reps are deeply invested in good service, and registration is a pressure point, so the tool had to reduce the rep's mental load in a moment that's already tense.

The core insight was that the interface should mirror what's actually happening between the rep and the guest, step for step. The rep sizes up who the person is, their member status, and how they're seeking entry; decides the best way to capture that—scan, swipe, or type; and builds, in effect, a profile of the guest's relationship to the club. So the app followed that exact sequence rather than the system's technical structure. Registration almost always comes down to getting a boarding pass into the system by one of three methods—scanning, swiping, or manual entry—and they aren't equal. Scanning is fastest and least error-prone; manual entry is slowest. I graded each method by what it requires, how easily it captures data, and how much training it demands, and let that grading set the interface's priorities: Scan became the primary path, Swipe second, manual methods available but deprioritized. I also flipped the language away from the technical backend toward the words that describe what the rep and guest are actually doing—Scan, Swipe—so a rep didn't have to learn complex registration types to use a quick tool. And because it's mobile with limited screen space, each flow showed only what's needed to complete that flow.