DSC Datathon
Self-hosted website for the 2025 Data Science Club Datathon, sponsored by the corner app and partnered with the NYU Stern Business Analytics Club.
CATEGORY
Frontend
NAME
Data Science Club Datathon Website
DATE
March - April 2025
Project link
LIVE WEBSITE
I built a comprehensive web application using Next.js to manage NYU's Data Science Club datathon event. My goal was to create a seamless platform where students could register, participate, and submit their projects. I chose a modern, dark-themed design that aligned with branding of the event sponsor, corner, to create an engaging user experience while ensuring the platform remained accessible and functional across all devices. This was the third design draft I had completed for the club (hence the "3" in the repository title), and we opted for this version due to its simplicity and functionality.
Registration System
I created a robust registration system that:
Handles team formations of up to 4 members
Validates NYU email addresses
Collects participants' experience levels in Python, ML, and Deep Learning
Uses react-hook-form for efficient form management
Integrates with Google Sheets for easy data management
Submission Portal
I developed a submission system where teams can:
Upload their Jupyter notebooks with commented code
Share their YouTube presentations (unlisted/private)
Submit presentation slides
Provide written reports
Select their challenge track (Easy or Hard)
Confirm compliance with Corner's data usage agreement
Space Management
I designed a clear way to display:
Event locations and their capacities
Study space availability
Room schedules
This helps participants plan their workspace and know where to be during different parts of the event.
Support Features
I included:
An FAQ section using shadcn's accordion component
Easy access to workshop materials
Direct links to Discord and email support
Optimize Spline Scene
I faced difficulties integrating the Spline scene into the landing page due to lagginess/bugginess of the animated scene on slower or mobile devices (and while I hate to admit it, sometimes the landing page doesn't work on Safari mobile because of this!) Using the Lighthouse tool in Chrome Developer Tools, as well as Spline's own optimization settings, I was able to significantly reduce the Largest Contentful Paint, Total Blocking Time and Speed Index by roughly 60%.
Form Complexity
One of my biggest challenges was managing the complex form state. With multiple required fields and conditional validation, I needed a robust solution. I chose react-hook-form which helped me:
Implement custom validation rules
Handle errors effectively
Reduce boilerplate code
Improve form performance
Data Storage
I needed a reliable way to store submission data. I decided to integrate with Google Sheets using Google Cloud Console because:
It's easily accessible for event organizers – I'm able to share the Google Sheet with anyone!
Provides real-time updates – The Google Sheet updates in realtime with new submissions from both forms.
Requires minimal maintenance and is relatively inexpensive (basically free with my NYU account)!
I implemented this using service account authentication and proper error handling.
User Experience Hurdles
Creating an intuitive interface for a multi-step process was tricky. I solved this by:
Organizing content into clear sections
Adding progress indicators
Providing immediate feedback on form validation
Maintaining consistent styling throughout
Using clear error messages
Security Concerns
I took security seriously and implemented:
NYU email validation
Data usage agreement confirmation
Secure submission endpoints
Environment variables for sensitive data
Responsive Design Challenges
Making everything work smoothly across devices was crucial. I:
Used Tailwind CSS for responsive layouts
Created custom breakpoints where needed
Designed with a mobile-first approach
Optimized typography and spacing for all screen sizes
The last two images show the other potential versions of the website for the datathon, and we eventually settled on the third version because it matched our sponsor's theme the most.
Building this platform was a challenging but rewarding experience. It pushed me to think carefully about user experience while maintaining security and performance. The platform successfully served its purpose for the datathon event, and I'm proud of how it turned out.
