Sync | Devpost

Inspiration

When our first break rolled around, we faced a familiar dilemma: pay nearly $100 for a solo Uber to the airport, or haul suitcases through multiple crowded subway transfers. While some students tried posting in group chats to find people to split rides, coordinating felt awkward, unreliable, and often fell through. With no secure or centralized system in place, most trips ended up being inefficient with multiple cars heading to the same destination, higher costs, and unnecessary traffic.

That’s why we built Sync. By matching students traveling at similar times and routes, Sync makes airport rides effortless to coordinate, dramatically cheaper, and more environmentally friendly. Fewer cars on the road means lower emissions, and a smoother, stress-free start to every break. No frantic messaging, no last-minute cancellations, just a ride that works for everyone.

What it does

Login: Users sign up with .columbia.edu emails via Supabase Auth.
Ride Requests: Users enter their pickup location, desired departure time, destination (e.g. airport or train station), and baggage size to create a ride request.
AI-Powered Compatibility: Our Snowflake Ride AI Optimizer analyzes historical ride data, timing flexibility, locations, and baggage constraints to match students and assign each match a clear compatibility score.
Coordination: Once matched, students unlock an in-app chat to coordinate exact pickup details, timing adjustments, pickup locations, and last-minute changes.
Smart Vehicle Recommendations: After a match is confirmed, the app recommends the optimal Uber or Lyft vehicle type based on group size and luggage to minimize cost.
Split Payments: After the ride is booked, users use Flowglad to automatically split and settle costs with one click — no Venmo chasing, no awkward math.

How we built it

Frontend: React, TypeScript, Vite, Tailwind CSS
Backend: Supabase for PostgreSQL database
AI: Snowflake and Gemini
Payments: Flowglad
Prototyping: Figma, Balsamiq

Challenges we ran into

Integrating Flowglad into the app, especially with customer-id-based transactions.
Displaying real-time updates such as chats and new ride requests.
Using tools for the first time, such as Snowflake Cortex and the Supabase database.

Accomplishments that we're proud of

Figuring out how to integrate tools that we had never used before, such as Flowglad, Snowflake Cortex, and Supabase.
Successfully creating a working app that has real-time updates and is able to communicate between multiple users and devices.

What we learned

It's important to ask for help when needed: when we were facing issues with implementing Flowglad with the customer-ids, we asked the team for help through their concierge service, and they were able to help us fix any issues.
Flexibility is key: when some aspects of our project weren't working, we decided to create new plans on the spot and pivot the sponsor tracks we were focusing on .

What's next for Sync

Expansion: We want to bring Sync to other universities and expand to bigger groups.
Advanced Analytics: Expanding our Snowflake usage to provide more insights.
Uber/Lyft API Integration: Moving from "recommending" a car type to allowing users to book Uber directly within the Sync interface.