You've sent the TikTok. You've made the spreadsheet. You've typed "bumping this" into the group chat for the fourth time. And somehow, the trip still isn't planned.
That's the gap droppit fills: one shared space where every idea, list, and plan actually lands, instead of disappearing into scroll-back. Let's take this trip out of the group chat.
Gen Z has trip inspo everywhere. The problem isn't finding ideas—it's keeping them organized.
Gen Z trip planning is very “save now, organize later.” Travel inspo starts on social—especially TikTok, Instagram, and Pinterest—but the actual planning still happens across screenshots, Notes, Google Sheets, and messy group chats.
TikTok travel videos get 40x more saves than comments, showing that people collect ideas faster than they act on them. A Troupe UX case study also found that 89% of people find group trip planning stressful, and 86% want trip details centralized in one place.
None fully combine inspo, docs, lists, sheets, and group visibility in one lightweight hub.
Instead of building a full itinerary app right away, the MVP focuses on the highest-impact behavior: making it easy to drop anything into one shared trip hub.
Create trip → Invite friends → Tap + → Add a doc → Everyone sees update
To keep the MVP focused, features like Voting (already a crowded space), Dedicated budget tools (handled through Sheets for now), Maps (useful later, but separate from core docs), and Rich text editing (kept lightweight for validation) are intentionally excluded.
Once the MVP was defined, I used Claude to shape the initial product prompts, then used Figma Make to build the prototype.
Claude helped me translate the research and MVP scope into clearer product logic, while Figma Make helped me move faster from idea to interface. I then refined the flow, screens, and product decisions myself to make the experience feel lightweight, social, and easy to use.
I’m happy to share that our team placed Top 3 🏆 in the “Best Design” category at the UW Women in Informatics 14th Annual Hackathon!
This year’s theme was “Depths of Discovery, Currents of Creation,” and my team and I prototyped an experience-first discovery app called Where You Go. We reimagined how people find places in a world where platforms like TikTok and Yelp often prioritize popularity & viral trends over what you’re actually looking for.
With Where You Go, users can search prompts like “Where should I go with my friends to grab a quick coffee & catch up?” and receive recommendations based on real lived experiences & intention, making exploration feel more authentic & personal.
I focused on designing the Onboarding flow 👩🏽💻 (introducing the concept in a familiar, intuitive way), the Home page experience 🏠 (with mood-based discovery & map exploration), and the Favorites page ❤️ (helping users save, filter, and return to meaningful places).
An experience-first discovery app designed around emotional landscapes, mood trends, and AI-powered recommendations.
Interactive walkthrough of the onboarding, home, search, and favorites flows.
We intentionally designed this to be educational about the app’s features by showcasing what exactly the app offers as users begin to create their account. This helps build initial trust and explains the experience-first discovery method.
The Home page was designed to be easy to use and provide multiple valuable insights all in one place:
The Search page uses AI to analyze and summarize community reflections, internet reviews, and social media content to extract emotions & generate cohesive "profiles" for each place. These profiles power personalized recommendations, trending experience categories, and our interactive city map.
All your saves are accessible in a clean, scrollable layout. Features additional filters and a search engine for easy viewing and quick filtering by mood.
Iterating rapidly under constraints while receiving invaluable mentorship from industry professionals.
Although the event moved incredibly fast, I gained so much insight from the mentorship we received throughout the weekend. Huge thank you to the designers from PitchBook & The Walt Disney Company who chatted with us — your guidance truly helped us solidify our concept and tell a stronger story.
Most importantly, shoutout to my amazing teammates! We spent hours ideating, designing, and building together, and this experience was genuinely so special because of your creativity & talent.
Explored how AI could help users diagnose and resolve device issues faster, while keeping IT administrators informed and in control of consequential decisions.
Windows 365 allows users to access a Cloud PC from supported devices. For organizations, this can simplify deployment and device management. However, when a device fails to connect, start, or function correctly, the support experience can become difficult for both the employee experiencing the issue and the IT administrator responsible for resolving it.
During an 8 week design sprint, I led exploration for how AI could reduce repetitive support work, help users understand what was happening, and carry useful diagnostic context into the troubleshooting process.
Users may encounter issues before they can sign in, connect to their Cloud PC, or access normal workplace support tools. They are then expected to identify and explain technical problems they may not understand.
For IT teams, incomplete support requests create additional investigation, repeated communication, and longer resolution times. Many common issues are straightforward, but administrators still need visibility into what the system is doing and control over actions with operational or financial consequences.
| The Friction | For Users | For IT Administrators |
|---|---|---|
| Accessibility | Support is difficult to reach from the affected device | Requests arrive through multiple fragmented channels |
| Context | Varying confidence levels being able to describe device issues | Tickets lack diagnostic details, requiring manual investigation |
| Resolution | Waiting for help creates downtime and frustration | Repetitive issues consume time, but automation feels risky without oversight |
I synthesized more than 40 insights from users, IT administrators, surveys, support-ticket patterns, and existing support experiences. Three themes consistently shaped the direction of the concept.
The proposed experience lets users access support from their phone, transfer relevant diagnostic information from the affected device, communicate through the format best suited to the situation, and involve an IT administrator when a consequential action is required.
The affected device could be frozen, disconnected, or unable to start. Designing the support experience only inside that device would make help unavailable at the moment it was needed most.
I used QR codes as a bridge between the computer and the user’s phone. A code placed on the physical device could provide a persistent support entry point, while an on-screen code could carry context from a specific error state.
A persistent support entry point allows users to begin even when they cannot sign in.
An error-state QR code can transfer issue context into the mobile experience.
Technical issues can be difficult to describe through text alone. Some problems are easier to show through a screenshot, while others benefit from a spoken conversation or step-by-step visual guidance.
The mobile experience supported text, voice, and images within the same troubleshooting journey. Users could begin with a suggested issue, describe something in their own words, share a screenshot, or switch to a call when needed.
Suggested issues and conversational troubleshooting
Users can share what they see instead of translating it into technical language
Users can move into a real-time guided conversation
AI could help diagnose common issues and complete repetitive troubleshooting steps, but not every action should happen automatically.
For decisions with greater operational or financial impact, such as replacing a device, the AI creates a concise record of completed diagnostics, identifies the likely issue, and recommends a next step. The administrator can then review the information and approve or reject the action.
The hardest part of designing AI support was not the conversational interface itself. It was coordinating the handoffs between a malfunctioning device, the user’s phone, the AI system, and the IT administrator.
This project taught me to think beyond individual screens and design the larger service experience: how users access help, how context moves between systems, where automation is appropriate, and when human judgment must remain involved.
Support products must remain understandable and accessible when the primary experience is unavailable.
Users and administrators need to understand what the AI knows, what it has done, and when it requires approval.
Switching devices should not mean restarting the problem or losing context.
This section contains a limited set of additional artifacts for portfolio conversations. The content has still been reviewed and simplified for confidentiality.
Design contributions for next-generation Windows experiences are currently protected by a Non-Disclosure Agreement.
For our senior capstone, my team partnered with U District Advocates to design and build a volunteer coordination platform supporting 30+ nonprofit organizations.
"You're more or less building a dating app—matching volunteers with organizations."
— Cory Crocker, President of U District AdvocatesInterviewed U District Advocates leadership and reviewed 5 volunteer platforms.
Research revealed three priorities that shaped the platform.
Surface upcoming events, volunteer activity, and key actions in one place.
Help organizations find, filter, and reconnect with volunteers beyond a single event.
Create, update, and communicate about events without relying on scattered tools.
We mapped the core organization flows, then used lightweight wireframes to validate structure before moving into high-fidelity design.
We built a functional frontend prototype using Lovable to rapidly validate our flows and test structure before investing time in high-fidelity Figma designs.
Open Prototype ↗
Surface upcoming events, volunteer activity, and key actions in one place.
Help organizations find, filter, and reconnect with volunteers beyond a single event.


Create, update, and communicate about events without relying on scattered tools.
Designed collaboratively across both teams; implementation led by our partner team.
Easily browse and find volunteer shifts that match personal interests and availability.


Keep track of upcoming commitments and seamlessly adjust shifts as needed.
View total hours contributed and reflect on the real-world community impact made.

Throughout the project, we continuously tested our assumptions through weekly team critiques, stakeholder reviews, and user feedback. Three moments in particular challenged our original direction and led to some of the project’s most meaningful design decisions.
Weekly sync w/ the team
Example of stakeholder feedback
Stakeholder and usability feedback moved directly into implementation, setting the foundation for launch.
I supported fast-moving Windows & Devices campaigns by jumping in wherever design support was needed, often working on quick-turnaround requests across multiple launches. I created and prepared visual assets for production while adapting quickly to shifting priorities and tight timelines.