Past the whiteboard: structured, editable flows built in days, not weeks.
Instead of static decks or free-form whiteboarding, I built each persona's flow as a rapidly-assembled interactive tool — drag steps between phases, double-click any card to edit, tag it MVP / integration point / open question. Built in Claude from interview notes and best-guess assumptions, then corrected live in the room. It turned alignment meetings into working sessions: NuCO2 and BarTrack could argue about a specific card instead of a vague impression, and open questions got tracked as data instead of getting lost after the call ended.
→ Open the interactive flow tool
Every touchpoint mapped against the system running underneath it.
A full service blueprint threaded the store manager's experience — onboarding through alert response through resolution — against three backstage lanes: the BarTrack platform, the "BarTrack system" lane (labeled explicitly as the sensors & AI detection engine reading pressure and pour data in real time), and NuCO2's own dispatch systems. Naming that sensor/AI layer explicitly — instead of leaving it implied — kept the team honest in reviews about what was actually automated versus what only looked automated from the front end.
→ Open the interactive blueprint
Regional visibility, designed to read as protection, not surveillance.
Corporate wanted fleet-wide uptime visibility; in stress-test sessions, store managers pushed back that real-time alerts to corporate would read as discipline, not support. The design resolution: a grace-period model — local resolution gets a window before anything escalates upward — paired with a regional view that frames downtime in dollars protected, not violations logged. The same reframe carried into alerting generally: NuCO2 dispatch gets every alert at the same moment the store manager does, so the store manager's screen had to answer a different question than "what's broken" — it had to answer "is this already being handled, and what do I need to do locally."
→ Open the live map — zoom in, click a store
A separate call: prototype in the real stack, not just in Figma.
Organized around how FOMs talk, not how the data was stored.
Forty-three candidate metrics came out of stakeholder sessions. Thirty were available without new engineering — that split became the MVP line. Within the 30, an IA exercise clustered them into groups a FOM would actually navigate by: Dealer Overview, Service Retention, Parts & Accessories, Operational Efficiency — and later, Marketing, added after FOMs asked for it.
Mapping three personas across a six-month arc.
The journey map tracked three personas at once — Inexperienced IC, Experienced IC, and Experienced Manager — from Post-Offer through Post-Core Onboarding. A manager's memory of "a great week 1" sat right next to a new hire's memory of "it took a while to be connected." Same event, two different stories.
Quick wins and bigger bets.
With the gaps mapped, we prioritized opportunities on an effort vs. impact matrix — separating what could move the needle right away from what was worth a bigger investment.
- Quick Wins: a standard welcome email before Day 1, a text from the manager or iBuddy, written 30/60/90-day expectations for each persona
- Bigger Bets: the onboarding portal itself, gamified checklists, automation of local dev-environment setup — the most friction-heavy part of week one
Threading a new model through flows people already trusted.
Eleven designers, four from Southwest's internal team. I led the group retrofitting seat selection into mWeb, iOS, and Android — designing for the full matrix of edge cases: lap children, unaccompanied minors, customers of size, companion passes, SWABIZ bookings. Every permutation organized into micro-journeys and accounted for before a line of code was written.
Scoping the work before a single screen was drawn.
A clear breakdown of MVP enhancements and ownership — entry points, bags updates, modify flows, responsive breakpoints — kept nothing falling through the cracks across eleven designers working in parallel.
When capacity got tight, we picked up testing ourselves.
The internal UX team owned Business Acceptance Testing — until capacity became a constraint and our team pivoted to run it directly alongside Southwest. Unglamorous, and exactly the kind of work that separates a project that ships clean from one that ships and needs damage control.
Live across web, iOS, and Android.
The redesigned experience shipped: a new seat selection model, a modernized booking funnel, and a consistent look and feel that honored the Southwest brand while moving it decisively forward.
Patient Transition Suite for naviHealth
Meet Mr. Ouch — your parent, your grandparent, someone tired and hurting who just wants to get home. It takes nurses, care coordinators, intake specialists, and physicians to get him there. Every one of them is a user of naviHealth's products, and the interface was serving none of them particularly well.
Card sorting, discovery workshops, and a role-based pitch nobody asked for.
The original engineering ask was a "lift and shift" — a visual facelift, no structural change. Research showed workflows varied a lot by role: a field nurse and a centralized data-entry specialist navigated the same screen to do fundamentally different jobs. That insight became the case for a role-based redesign instead — a conversation that changed how UX was valued on the project.
Reorganizing around the F-Pattern, not the database.
The legacy layout forced an inefficient, non-linear scan just to complete basic tasks. Mapping the data users needed most against the F-Pattern — the natural eye-tracking path people take through dense interfaces — put the right data in the right place and cut cognitive load.
Staying lo-fi on purpose.
Low-fidelity wireframes kept users focused on layout and interaction, not visual polish still being developed in parallel — letting people give honest feedback on structure. Testing surfaced real efficiencies, like re-sequencing tasks to cut extraneous clicks — observations from the people doing the work, not designer opinions.
Reorganized around clinical workflow, not the database.
31 interviews. Surveys from 70+ physicians and nurses. Seven personas.
A team of six designers and researchers ran 31 one-on-one interviews with potential patients, reviewed survey responses from more than 70 physicians and nurses, and gathered input from a dozen business leaders — all within a few weeks. Clustering and affinity mapping produced seven grounded personas that gave the whole team a shared language.
The most significant finding was one nobody expected.
An analytics deep-dive with HCA's Marketing Services team mapped what people were actually looking for and where they dropped off. A substantial share of traffic was people hunting for career content — buried several levels deep. They were finding it despite the site, not because of it. That reframed the whole priority list.
A shared vocabulary, built in the open.
I facilitated a four-hour workshop walking stakeholders through all seven personas, running How Might We exercises, then live-sorting ideas with group dot-voting.
Three directions, presented to leadership for input.
Grounded in the research and guided by the personas, the team produced three distinct directions — low-fi wireframes through mid-fidelity visual explorations — before landing on a serious modernization that stayed inside HCA's style guide.
Modernized, without losing HCA's warmth.
A modernized homepage balancing patient, provider, and investor needs within HCA's existing style guide — a serious step forward that still feels unmistakably HCA.
