OpenSeat

Inspiration

As MBA exchange students arriving in Chapel Hill from Latin America, we faced a combination of challenges that many newcomers know too well: no car, no local network, and a campus where getting around without a vehicle is genuinely difficult.Existing carpooling apps solve the logistics — but not the trust. Getting into a stranger's car is uncomfortable. We wanted a solution that felt safe, familiar, and social. Knowing that the person in the driver's seat goes to your same university, works at your same company, or belongs to your same club changes everything. That experience inspired us to build OpenSeat.

What it does

OpenSeat is a mobile-first web app for carpooling within closed communities — universities, companies, and clubs. Users verify their identity through their institutional email (such as @unc.edu), ensuring every person on the platform belongs to the same community. From the home screen, users can:

Offer or find rides with origin, destination, departure window, seat availability, and vibe preference
Browse upcoming community events and see who's going — both drivers and passengers
Join a ride to an event in one tap, with the destination pre-loaded
Choose their pickup preference — whether the driver comes to them, they walk to the driver, or they meet halfway Our matching algorithm surfaces social hints — shared majors, interests, and clubs — so you're not just sharing a car, you're meeting someone you already have something in common with. After each ride, users rate each other with a Good Vibes score and can choose to connect, building a social graph within their community. Every ride also tracks CO₂ saved, contributing to a measurable environmental impact.

How we built it

We built OpenSeat using:

Next.js 14 with TypeScript and Tailwind CSS for a mobile-first frontend
Firebase for authentication, real-time Firestore database, and file storage
- Geocoding API (address conversion), Google Maps JavaScript API (map rendering), Directions API (route calculation) and Places API for routing and address autocomplete
Vercel for automatic deployment from GitHub
Claude Code as our AI pair programmer across four parallel workstreams

Challenges we ran into

Our biggest challenge was coordination. Four people building in parallel means four people who can block each other. We solved this with a mock-first development strategy — all data functions were designed with their final interface but returned hardcoded mock data initially, so the UI team could work independently of the backend.

Accomplishments that we're proud of

Building a fully functional app from scratch in one day — as a team of four people coming from completely different backgrounds. None of us are software engineers by training. We're MBA students who came together around a shared problem, and turned it into a working product. Every idea made it in. The trust and safety angle, the social layer, the community events feature, the Good Vibes score — all of them came from real conversations the four of us had about what we wished had existed when we arrived in Chapel Hill. The fact that we could take those ideas and see them live on a phone screen by the end of the day is something we're genuinely proud of.

What we learned

We learned that in a hackathon, coordination is the real bottleneck — not code speed. Investing the first hour in documentation paid off enormously: every team member and AI tool had the same source of truth. We also learned to ruthlessly prioritize a perfect demo over many features — one flow that works flawlessly beats ten features that half-work.

What's next for OpenSeat

Real-time in-app chat
Recurring rides for daily commuters
Venmo and Zelle integration to split gas costs
Push notifications for ride requests and event reminders
Expansion to other universities, corporate campuses, and clubs
A B2B SaaS model where organizations onboard their communities in minutes — from Chapel Hill to Buenos Aires