Problem
City Tech is a commuter school. The ride to campus is long, often without signal, and mostly spent scrolling, sleeping or stressing. It's a big part of student life that nobody designs for.
← All work
2026AI Design Hackathon
Case study 03 · 24-hour hackathon
Commute Mode: turning the subway ride into student time.
City Tech students spend 30 to 90 minutes on the train every day, and nobody designs for that time. At the AI Design Hackathon, three of us went from brief to pitch in under 24 hours, and won 2nd place.
🏆 2nd place
⏱ Under 24 hours, brief to pitch
3 designers
The 60-second version
At a glance
What I did
- Designed every screen of the Commute Mode app
- Made the whole 50-second pitch film: AI images, animation, voiceover and edit
- Interviewed City Tech commuters with the team
- Worked through the pivot together, from a citywide system to one small touchpoint
Outcome
- 2nd place at the AI Design Hackathon
- A QR touchpoint in train cars that opens a personal ride plan
- Built, rushed and pitched in under 24 hours
01 · The brief
Design the space between.
The challenge came from City Tech's Communication Design department and Designpreneurs Hackathon. Talk to real students, focus on one problem, and use AI to sharpen the thinking, not replace it. The brief was just as clear on what not to do: don't try to solve the entire MTA, and don't build a generic chatbot.
The prompt
How might we redesign the City Tech commute so the time between home and campus works for students, not against them?
How the judges scored it
Creative concept & idea40%
Execution & craft20%
Presentation & story20%
Problem, insight & framing15%
X-factor5%
The idea was worth more than the polish.
The concept counted for 40% of the score, and the brief reminded us that subway riders give you seconds, not minutes. So we spent our limited hours on one clear idea a stranger could get at a glance.
02 · The 24 hours
Brief to pitch, fast.
01
The brief
Make the commute work for City Tech students.
02
Talk to riders
Quick interviews with real City Tech commuters.
03
Idea one
A citywide system of kiosks and hubs.
04
The pivot
Smaller, offline-first, inside the train car.
05
Make it real
Screens, AI storyboard, AI film and voiceover.
06
Pitch
Presented to the judges. 2nd place.
Everything on this page was made in under 24 hours, and it shows in places. The screens are first passes, the research was a handful of quick conversations, and the film was cut against the clock. What we optimized for was one clear idea, grounded in real students, that judges could understand in seconds.
03 · Research
We asked commuters what the ride is really like.
Following the brief, we spent the first part of the hackathon talking to City Tech students. Three questions: what do you do on your commute, what would you actually use, and would you rather relax or be productive?
"Relax, because the day is packed with schoolwork."On relax vs. productive
"Be productive, whether that's for work or school."On relax vs. productive
"Listen to music, sometimes read class notes."On what they do now
"Study, or play a game with flashcards."On what they'd use
"When I lose service on the train, there's no way for me to feel safe."The quote that changed our concept
Insight 1
The ride is mostly passive. Music, scrolling, sleeping, waiting. The time is there, it just isn't designed for.
Insight 2
Every day needs something different. Some students want to switch off. Others want to get ahead before an exam.
Insight 3
It's emotional. Commuting can feel draining and isolating, and music or sleep is how many students cope.
Insight 4
No signal underground. Any digital tool for the subway has to work offline.
04 · The pivot
Our first idea was too big.
Idea one
A citywide commuter system
- Kiosks and support hubs across the city
- Tap in with your Apple Wallet student ID
- Connected to CUNYfirst and Brightspace
- Alerts from home to the station exit
What we pitched
Commute Mode, inside the train car
- Permanent QR touchpoints by the doors and route maps
- Sync before you go, so everything works with no signal
- Three Ride Modes: Relax, Prepare, Productive
- Anonymous signals that you're not riding alone
Two things made us cut it down.
A student told us they lose service on the train, which a network of kiosks couldn't fix. And the brief warned against trying to solve the entire MTA. A QR code in a train car plus an offline sync is small enough to actually deploy, and it answers the problem students really described.
05 · How it works
One commute, four moments.
01 · Before you leave
Sync before you go.
One tap downloads notes, flashcards, schedules and playlists from the apps students already use. Everything works underground with no signal: our direct answer to "I lose service on the train."
02 · Heading to the station
It notices you're commuting.
Near the station, a gentle prompt asks if you want to start Commute Mode. No app to remember to open at 8 a.m.
03 · On the train
Scan once, pick a Ride Mode.
A QR code by the door opens the session. Relax, Prepare or Productive, each one sized by AI to how long your ride is. Anonymous signals like "6 City Tech students are on this train" make the ride feel less lonely, without showing anyone's name.
04 · Jay Street–MetroTech
Arrive ready, not drained.
Same train, same 45 minutes. But this time you walk into class having reset, rehearsed or reviewed, instead of scrolling.
06 · Try it
Plan your own ride.
Where AI actually helps: not as a chatbot, but in timing. A 15-minute ride gets quick flashcards or a breathing reset. A 90-minute ride gets a real study sprint. A presentation on your calendar pushes Prepare to the top. Midterm week changes what's recommended.
07 · The screens
First passes, made in hours.
My first part. I designed every screen of the app: onboarding, the ride-length picker, Sync before you go, the station prompt and all three Ride Modes. A calm, dark interface made for a crowded train at 8 a.m.: one decision per screen, big tap targets, nothing to read twice.
08 · Making the film
No camera crew. No actors. A few hours.
My second part: I made the whole film. A pitch lives or dies on the story, and there was no time to film anything, so AI became my production crew. I generated the storyboard as still frames, animated each one into a short video shot, generated the voiceover, and edited it all into a 50-second film with the app screens layered on top. Every step, in a few hours.
Storyboard stills (Gemini)→AI video shots→Voiceover (ElevenLabs)→Edit + app screens→50-second film
AI doesn't get it right the first time.
Take 1
Final
09 · Result
2nd place, and a team I'd work with again.
We pitched Commute Mode to the judges with the film, the screens and the research behind them, and took 2nd place at the AI Design Hackathon, run by City Tech's Communication Design department with Designpreneurs Hackathon.
Not bad for an idea that didn't exist the day before.
10 · Reflection
What 24 hours taught me.
What worked
- Research first, even in a rush. One quote about losing signal reshaped the whole concept.
- Shrinking the idea. A QR code in a train car beat a citywide system we couldn't explain in a minute.
- AI for production, people for judgment. AI made the film possible. Deciding what it should say was still on us.
With more time
- Test the QR placement and scan flow on a real train, during rush hour.
- Run Ride Modes with a few commuters for a week before polishing any UI.
- Design the privacy and accessibility side of the community signals properly.
- Take the screens past first drafts.