UX Design
Wanted UX-related experience and expected to find UX Design represented among the available interests.
LaptopI've been the frustrated user. I've also been the person helping one. Years of running my own business and working in customer experience showed me where systems break down for people. My psychology degree helps me understand why, and my UX design degree helps me design something even better.
A UX designer has to understand the user, just like a hiring manager needs to understand their future employee.
Thanks for getting to know me.
Now that you know me, see how I work.
Designing around what students actually need.
How might we make discovering and exploring experiential learning opportunities on PeopleGrove more intuitive for students?
02 / RESEARCH
Individual interviews, the same questions every time.
Wanted UX-related experience and expected to find UX Design represented among the available interests.
LaptopInterested in fashion, beauty, and digital/social media. Wanted opportunities aligned with her specific marketing interests.
LaptopPrimarily accessed PeopleGrove by phone and wanted to explore opportunities efficiently.
Phone03 / FINDINGS
These are screenshots of the live PeopleGrove site on a phone. They are not my design. They show what students see now, and what the three findings below are based on. My designs begin on the Wireframes page.
EXHIBIT A
EXHIBIT B
EXHIBIT C
EXHIBIT D
“UX Design” and “UX” returned no match, while “Design” surfaced Graphic Design instead.
“I just didn’t select anything. I didn’t want to choose something that wasn’t actually my interest.”
P1
Enable students to express specific interests even when an exact predefined category is unavailable.
A broad “Marketing” category did not reflect the type of marketing P2 wanted to pursue.
“Just because something says marketing doesn’t mean it’s the type of marketing I’m interested in.”
P2
Support refinement by industry, specialization, opportunity type, and work preference.
P3 could finish tasks on her phone, but it took extra scrolling and effort.
“I want it to feel like it was actually made for a phone instead of like I’m looking at the computer version on a smaller screen.”
P3
Prioritize essential information and controls on smaller screens so students can complete core tasks with less scrolling and navigation.
SAMPLE NOTEEach barrier emerged most clearly from one participant. The next round of research would test how broadly these patterns appear across a larger group.
WHAT ALREADY WORKEDSearch was easy to understand, and saving helped students return to opportunities.Preserve both.
04 / STUDENT PERSONA
Jordan is a composite of the behaviors and frustrations observed across the interviews, not a single participant.
SYNTHESIZED PERSONA
Student seeking relevant experiential learning opportunities
“Find relevant opportunities that reflect my specific interests, without unnecessary effort.”
Constructed persona statementACCESSLaptop and phone
EMPATHY MAP / JORDAN
“I know what I’m looking for. I need a way to express it here.”
Illustrative synthesized statementDoes this reflect my interests? Is this opportunity worth reviewing?
Searches, refines results, reviews opportunities, and saves promising options.
Frustrated by irrelevant results and extra effort. More confident when tasks feel straightforward.
WHERE THE FRICTION SHOWED UP
Synthesized from three participant perspectives. This journey does not represent a single observed session. Pink marks the stages with the most friction.
THE KEY OPPORTUNITYHelp students express their interests, discover opportunities that reflect their preferences, and evaluate them efficiently across devices.
05 / WIREFRAMES
Six proposed mobile screens, informed by student research.
Click the wireframe to enlargeTap the wireframe to enlarge
Quick access to search and key actions
REQUIREMENTS ADDRESSED
DESIGN DECISIONS
Start a new search and explore options
REQUIREMENTS ADDRESSED
DESIGN DECISIONS
Narrow results based on specific needs
REQUIREMENTS ADDRESSED
DESIGN DECISIONS
Scan and compare opportunities
REQUIREMENTS ADDRESSED
DESIGN DECISIONS
View full information and take action
REQUIREMENTS ADDRESSED
DESIGN DECISIONS
Add and edit interests with flexibility
REQUIREMENTS ADDRESSED
DESIGN DECISIONS
06 / EXPERIENCE
Tap through the concept, or follow the guide. Each screen explains the decision behind it.
Sample opportunities shown for demonstration.
07 / REFLECTION
Three recurring barriers shaped the design direction: expressing specific interests, finding relevant opportunities, and navigating efficiently on mobile.
Five students, three tasks on the clickable prototype: find an opportunity in their field, refine it by a specific interest, and save one.
A research-informed proposal, not a validated solution.
The six wireframes and the clickable prototype have not undergone usability testing.
Assumes PeopleGrove can offer interest suggestions and custom interests on its existing platform. Engineering would need to confirm how new interests map to current categories and search.
Touch targets and contrast sized for a phone, labels for screen readers on filters and the save action, and no gestures that need precision.
“Improving discovery isn’t always about adding more options. Sometimes it’s about giving people a better way to communicate what they’re actually looking for.”
END OF PEOPLEGROVE
Thanks for reading.
Experience
Where human behavior, operational realities, and thoughtful design come together.
Each chapter highlights a different part of my journey and shows how my background and experiences inform the work I do today.
A foundation in human behavior that shapes how I approach problems and design solutions.
Formal education and continuous learning that strengthened my UX mindset and design capabilities.
Applying research and design thinking in a global airline environment.
Contributing to mission driven product through user research and strategic design.
The tools, methods, and mindsets I rely on to turn insights into meaningful solutions.
Bridging user needs and business goals to build solutions that create lasting value.
Thanks for visiting. This marks the end of my portfolio, but I’d love to keep the conversation going.
Résumé
A concise version of the experience behind the work.
Work
Three case studies: Inside the Club, PeopleGrove, and Blend Sessions.
EXPERIENCE CONCEPT
Reshaping the premium experience during irregular operations.
Independent concept. Not affiliated with or endorsed by Delta Air Lines. Delta marks are used for illustration only.
02 / THE PROBLEM
Sky Club guests enter the lounge for a calmer, more comfortable travel experience. But when a flight disruption requires personalized help, that experience can quickly break down.
Guests may need to leave the lounge, return to crowded gate areas, or rely on phone support to get help, extending the disruption beyond the flight itself.
“The problem was not
the disruption.
It was what happened next.”
TODAY: THE CURRENT-STATE JOURNEY
Receives a cancellation alert while relaxing in the Sky Club.
Seeks help at the Sky Club front desk and is directed to the gate agent.
Waits in line to speak with a gate agent about rebooking.
Goes back to the Sky Club and waits in another long line to re-enter.
Sky Club guests experiencing flight disruptions need a way to access personalized assistance without abandoning the environment they entered the lounge to receive.
How might we help Sky Club guests request, monitor, and receive disruption assistance while allowing them to remain inside the lounge?
A premium lounge promises calm. A disruption is exactly when guests need that promise to hold.
03 / DISCOVERY
I conducted primary research with 21 participants across multiple hub airports, including Delta SkyMiles Reserve Card Members and Diamond Medallion members, during irregular operations. Three signals emerged.
Details are anonymized, and no personal information is shown.
Expressed a need for better support options within the Sky Club.
Reported leaving the lounge during active disruptions to seek assistance.
Showed sentiment recovery after their issue was resolved.
These are findings from my participant sample, not claims about all Sky Club customers.
“I’m currently sitting in the Sky Club. I wish there was an agent here who could assist with my flight.”Participant 1, Diamond
“Having to call in during every irregular operation is inefficient. Agents often redirect you to the gate.”Participant 2, Diamond
“Delta used to have a help desk in the Sky Club. It would make a significant difference.”Participant 3, Diamond 3MM
WHAT THE RESEARCH POINTED TO
Guests wanted assistance without leaving the lounge.
Travelers needed to know whether help was available and when they would receive it.
The need was not simply another self-service tool. Some situations still required a conversation with a person.
Support should solve the travel problem without creating a second disruption inside the lounge.
WHO THIS DESCRIBES
A Diamond-level frequent flyer, in the lounge when plans change.
Synthesized from recurring themes across the research, not a single participant.
GOALResolve the disruption without leaving the lounge.
PAIN POINTSEMPATHY MAP
“I wish there were an agent here who could help with my flight.”
“Do I really have to leave the lounge to get help? How long is this going to take?”
Leaves the lounge, calls support, or gets redirected to the gate for assistance.
Frustrated and uncertain while trying to get help. Relieved once it’s resolved.
04 / DESIGN DIRECTION
Guests should be able to request and wait for help without leaving the Sky Club.
AccessStarting a request should take less than two minutes.
PreservationGuests should not have to stand in another physical line inside the lounge.
PreservationGuests should know their request was received, where they are in the queue, and when help is approaching.
UncertaintyRequests tied to imminent departures or missed connections should be surfaced first.
UncertaintyTechnology should organize access to an agent, not replace the conversation.
Human supportTHE SERVICE CONCEPT
WHAT WOULD HAVE TO BE TRUE
Assumptions to confirm with operations and engineering.
Who the kiosk has to work for.
Concept stage: these are requirements to validate, not tested results.
05 / THE FLOW
THE GUEST JOURNEY
What the guest sees
BEHIND THE SCENES: THE SERVICE BLUEPRINT
Swipe sideways to see all four stages.
06 / EXPLORATION
From early concepts to a refined interaction, the design evolved to help Sky Club guests request flight assistance quickly, quietly, and on their own terms.
Follow the wireframe journey
Select a fidelity level, then step through the seven screens.
WHAT CHANGEDMapped the flow with placeholder boxes and notes, and annotated why each screen exists.
WHAT CHANGEDMoved from placeholder boxes and notes to real copy, icons, and the lounge setting.
WHAT CHANGEDApplied Delta’s visual system: navy and red, real flight data, and clear status states.
Copy and sample flight data changed between rounds as the design evolved.
THREE TRADEOFFS THAT SHAPED THE EXPERIENCE
First-come, first-served could disadvantage guests with imminent departures.
Decision. Use live flight data to prioritize time-sensitive requests rather than asking guests to self-declare urgency.
Automation can improve efficiency, but complex disruptions still require human judgment.
Decision. Use the kiosk to organize access to an agent, not replace the agent.
Guests need to know their status without exposing personal travel information.
Decision. Use an anonymous queue ID, such as A17, across shared lounge displays.
07 / THE EXPERIENCE
Walk through the four moments of the proposed experience, from requesting assistance to meeting with an agent.
Sample flight data shown for demonstration.
Need help with
a flight issue?
Get connected to a Sky Club agent without leaving the lounge.
Select all that apply.
Enter your SkyMiles number and name to see today’s trip.
Enter flight details manually
Show more flights for today
Optional
0/300
Returning to home in 15s
9:41
Concept interface shown for demonstration.
08 / REFLECTION
Separating the disruption itself from what happened afterward changed the direction of the solution.
Usability testing is next.
Five travelers, three tasks on the prototype: request assistance, join the queue, and find their way to an agent.
WHAT I WOULD TEST
Proposed measures, not results.
A premium experience should not disappear the moment travel stops going according to plan.
END OF INSIDE THE CLUB
Thanks for reading.
PRIVATE · TRACK SHEET
| Case study | People | Opens | Came back | From index | Finished | Furthest page | Used prototype | Next project |
|---|
INDEPENDENT UX CONCEPT
Listening sessions evolve. Playlists do not.
A more fluid way to move between the playlists that already define how you listen.
Pick 2–4 playlists.
Influence changes the order, not your playlists.
Uses your selected playlists to guide additional picks. Your original playlists stay unchanged.
Active blend
Late Night Blend is now available in your library.
The session will end without changing your original playlists.
Interactive. Pick two playlists to start a blend.
Sample content shown for demonstration.
02 / RESEARCH
People build playlists with distinct identities, but move between them manually.
THE PATTERN
People organize music into separate moods, genres, and contexts, then switch between them manually when the moment changes.
RESEARCH
LIMITATIONFive participants is a small sample, so these findings indicate patterns, not proof.
Blend Sessions is introduced only after the seven questions, allowing existing behaviors to surface before participants react to the concept.
OPTIONAL CONCEPT CHECKAfter the interview, show the concept and ask: “Looking at this concept, what do you think it would let you do?”
MEET THE FIVE LISTENERS
Chooses playlists around the moment: getting ready, driving, working, or winding down.
“I’d want to tell it which playlists I’m in the mood for and have it move between them for me.”
Opens recently played music or a recommended mix and lets it run.
“Give me variety, but keep it inside the music I already picked.”
Keeps many highly specific playlists and chooses based on mood.
“Let me mix them for the moment without actually combining them.”
Chooses music based on who is around, especially while driving or hosting.
“I want to set the vibe once and stop DJing my own playlist all night.”
Knows exactly which playlists he wants and does not use recommendations heavily.
“If I pick four playlists, I’d want some control over which one shows up more.”
THREE FINDINGS
Listeners move between playlists as the mood, activity, or setting changes, and manage those transitions manually.
P1 went from an upbeat playlist to a calmer one on the way to the car. P4 started queueing songs by hand when one playlist got too slow.
“I want to set the vibe once and stop DJing my own playlist all night.” P4
DESIGN IMPLICATION
Reduce manual switching by blending several playlist sources into one continuous session.
Raised by P1, P2, P3, P4, P5
Participants wanted movement across their own playlists without handing the session to recommendations.
P1 avoids generated mixes that bring in music she does not want. P2 went back to a focus playlist when a generated mix got too energetic. P5 wants control over which playlist shows up more.
“Give me variety, but keep it inside the music I already picked.” P2
DESIGN IMPLICATION
Let the listener choose the sources while the system handles the sequencing.
Raised by P1, P2, P5
Listeners wanted playlists to work together without permanently changing the collections they had built.
P3 rarely combines playlists because merging creates clutter and changes collections she spent time curating. P2 keeps separate instrumental, jazz, and low-key playlists.
Supporting behavior: P4 manually queued songs across playlists, while P5 described building temporary queues rather than permanently combining collections.
“Let me mix them for the moment without actually combining them.” P3
DESIGN IMPLICATION
Blend at the session level, keep the originals unchanged, and make saving optional.
Raised by P2, P3. Supporting behavior from P4, P5.
WHAT DIDN’T FIT
Not every participant experienced the problem in the same way. P2 was comfortable letting recommendations run and rarely managed queues manually, making this behavior less relevant to their listening habits. P5, however, wanted more control over how strongly each playlist influenced the session, which reinforced the need for adjustable influence rather than a fully automated blend.
03 / DIRECTION
Each one traces back to what the five listeners described.
The listener decides whether a blend is saved or discarded.
FROM FINDING 03
The listener chooses the playlists that define the blend. Any added music should stay within those selected boundaries.
FROM FINDING 02
Playlist influence shapes the sequence, so the listener does not have to build a queue manually.
FROM FINDINGS 01 AND 02
The blend does not change the playlists it draws from.
FROM FINDING 03
HOW THIS DIFFERS
Blend Sessions starts with the playlists already in the listener’s library, using them to define the sound and direction of the session. The system sequences across those sources and can introduce fitting music without changing the original playlists.
The listener sets the boundaries. The system shapes the flow.
04 / EXPLORATION
Current behavior comes from the five interviews. The blend flow is the proposed design.
BEFORE BLEND SESSIONS
Composite of behaviors described across the five interviews.
WITH BLEND SESSIONS
DESIGN HYPOTHESIS
The design moves playlist management to the beginning of the session, so listeners spend less time managing music while it plays.
THE THREE SCREENS
Pick 2–4 playlists.
Influence changes the order, not your playlists.
Uses your selected playlists to guide additional picks. Your original playlists stay unchanged.
Active blend
EDGE STATES
FOUR TRADEOFFS THAT SHAPED THE EXPERIENCE
CONTROL VS. AUTOMATION
Too much control creates work. Too much automation removes ownership.
Decision. Let the listener define the sources and influence while the system handles sequencing.
IDENTITY VS. DISCOVERY
Blending existing playlists should preserve the listener’s taste without closing off discovery.
Decision. Start with playlists the listener already created, then allow fitting music that stays within the blend’s boundaries.
VARIATION VS. MOOD STABILITY
Mixing playlists can feel jarring when transitions shift too abruptly.
Decision. Keep the sequence smooth and make playlist shifts visible.
TEMPORARY VS. PERMANENT
Listeners may want to explore without committing.
Decision. Keep blends temporary unless the listener chooses to save them.
05 / THE EXPERIENCE
Choose playlists, set the influence, and listen. Nothing is saved unless you choose to save it.
Sample content shown for demonstration.
Pick 2–4 playlists.
Influence changes the order, not your playlists.
Uses your selected playlists to guide additional picks. Your original playlists stay unchanged.
Active blend
Late Night Blend is now available in your library.
The session will end without changing your original playlists.
06 / REFLECTION
Listeners choose the playlists, set their influence, decide whether discovery is included, and choose whether to save. The system handles sequencing inside those boundaries.
A concept, not a validated solution.
Five discovery interviews informed the direction. The interactive prototype is ready for usability testing.
Five listeners, three tasks on the prototype: set up a blend, adjust the influence, and decide whether to save it.
WHAT I WOULD TEST
FEASIBILITY
Assumes the app can sequence tracks across several playlists and label fitting picks as it goes. The rules for fitting music would need a recommendation team to define and test.
ACCESSIBILITY
Each influence slider needs a text value and a non-slider way to adjust it, selected playlists must be marked by more than color, and the fitting-music switch needs a clear screen reader label.
PROPOSED MEASURES, NOT RESULTS
People do not always want to choose between their playlists. Sometimes they want all of them, at once, for as long as the moment lasts.
END OF BLEND SESSIONS
Thanks for listening.