Weekend in Harlem
Jazz history, local bakeries, and a sunset walk through St. Nicholas Park.
Graduate studio · 2025
Navi
The heatmap could show people where to go, but not what to do when they got there. Learn, Plan, Go grew out of that gap.
- Role
- UI/UX Designer
- Scope
- I synthesized 14 resident and stakeholder responses, including two Manhattan businesses, adjusted the information architecture, and created the design system.
- Outcome
- A functional team Figma prototype, followed by my solo React portfolio demo.
- Timeline
- January 2025 – June 2025
- Tools
- Figma, FigJam, React, TypeScript
Key moves
- Tested: Other design students reviewed the homepage and search. We aligned card heights and cut copy and tags.
- Untested: The semester ended before Learn, Plan, Go or booking could be tested.
- Current demo: Plan's full cost breakdown and Go's routing are not built.
An early heatmap showed where to go
The brief asked how regenerative tourism could be participatory and contributive for visitors, local businesses and artisans, and longtime residents. The Manhattan heatmap was our first idea, not live tourist-density data.
The first prototype: a Manhattan heatmap
The regions reflect early emphasis. We had no live tourist-density or geo-analytics data.
Select from the list to preview how Learn would frame each of 12 neighborhoods.
People wanted to know what a neighborhood had to offer
Platform audits and resident research
Interviews were hard to recruit, so surveys became the primary method. I collected 14 resident and stakeholder responses, including two Manhattan businesses, and spoke with NYC Tourism. The sample included longtime residents and many college-aged people, so it doesn't stand in for all NYC residents.
The team audited six travel platforms. I evaluated Airbnb with Kaori Ogawa and Amy Zhang against Nielsen's ten heuristics.
Guest Favorite label overuse
Nearly every listing we reviewed carried the badge, but we couldn't see consistent criteria. That made it less useful as a quick decision signal.
Minimal family & accessibility filters
Kid-friendly and accessibility filters were thin. A crib was often the only family-facing signal, so group planners had little to compare.
Visual clutter and repetitive listings
Similar photography and copy made listings hard to distinguish before a traveler could compare neighborhood context.
Two survey concerns appeared most often.
10 of 14 responses (71%) named overcrowding and over-tourism. 7 of 14 (50%) named a lack of authentic experiences.
Survey responses also raised rising costs, local-business displacement, repeat visitor relationships, and discovery outside traditional social platforms. Tourism professionals described regenerative work already happening citywide. From what I remember, people wanted deeper engagement with a neighborhood and its offerings.
Highlights from 14 responses, including 2 local businesses: Reimagining NYC Tourism: A More Meaningful & Sustainable Experience (resident and stakeholder survey).
71%
50%
How Learn, Plan, Go addressed the research
I grouped responses into three archetypes: digital nomads, ethical travelers, and visitors already contributing to communities. These were early audience hypotheses, not validated market segments.
I used Learn, Plan, Go to bridge customers and businesses:
- Learn covers the neighborhood, activity, and host so someone can try something new without feeling like a fish out of water.
- Plan is supposed to help someone, well, plan. It shows activity requirements and a small-business cost breakdown.
- Goshows eco-friendly ways to reach the event from the visitor's current location.
In the prototype, visitors could choose a date and book directly with a partner in a separate module beside those stages.
Cost, location, and event type were non-negotiable. Time-sensitive statuses came first. Going Fast reflected remaining availability. Popular was meant to compare tickets booked within a set window against other events. I didn't wire that logic into React. In another course, my team noticed Airbnb used Featured so often that it lost value. I kept that in mind here.
Research and product scope
Each row connects a finding to the product area it affected.
Cain
- Need
- Coordinate an activity around a group's needs.
- Product area
- Group planning
- Scope
- Future opportunity
Ororo
- Need
- See what is happening in neighborhoods nearby.
- Product area
- Neighborhood exploration
- Scope
- Implemented
Selina
- Need
- Compare a shorter trip with precise filters.
- Product area
- Search and filters
- Scope
- Implemented
View the journey and booking artifacts
Awareness
- Traveler moment
- Look beyond the same tourist-heavy areas.
- Product need
- Nearby neighborhood context
Consideration
- Traveler moment
- Compare an activity, its host, timing, and requirements.
- Product need
- Details repeated at decision points
Decision
- Traveler moment
- Review the full cost before confirming.
- Product need
- Transparent booking summary
Neighborhood discovery
Start with a local area.
Activity detail
Review the host, activity, and requirements.
Date and time
Choose an available session.
Cost review
See the total before committing.
Confirmation
Keep the activity details and schedule together.
I rebuilt the booking flow in React
Rebuilding Navi as a working system
We didn't have engineering resources or a budget, so the project stayed in Figma. My teammate proposed the original information architecture. I tweaked it to match the personas and journey maps. The team had trouble working together. Our experience with design tools and schedules didn't line up. I created the design system from head to toe, saving time on aesthetic decisions and refocusing us on resident and business-owner insights. The journeys and flows were internal planning artifacts with no engineering handoff.
The homepage was sort of a first pass at whether the system held up. Other design students tested the homepage and search, so it makes sense that feedback centered on layout. We matched heights, then cut tags and copy to lessen visual load.
By the time the semester ended, we didn't have an opportunity to test Learn, Plan, Go or the booking flow. The functional Figma prototype included events that felt native to each neighborhood, a booking widget, and a profile of the event host. I had plans for group event coordination, but I was comfortable with that as a V1.
After the semester, I rebuilt the system in React and TypeScript on my own. The current version covers the individual booking flow. Plan doesn't show the full cost breakdown yet. Go lists public transit, walking, and bike options for each activity, but it doesn't calculate a route from the visitor's current location yet.
The system page and booking demo share components.
Editorial homepage slice
Plan with the neighborhood in view
Locally ownedNature first
Chinatown food crawl
Dim sum counters and tea shops, with context beyond star ratings.
LES art walk
Gallery openings and community murals, timed for quieter weekday mornings.
A working booking flow
What works now and what still needs testing
I could see NYC Tourism as a future partner, but that partnership hasn't happened. The browser demo supports the individual flow. I still need to test it with residents, travelers, and local hosts.
Working now
- Live component system and editable variants
- Neighborhood exploration and filters
- Working individual booking flow
Next research
- Deeper Learn pages
- Local host and business onboarding
- Group booking remains a future opportunity and is not wired into this demo.
