Pilot Episode | Devpost

Inspiration

As Computer Science students at UW–Madison, coding and projects aren’t just assignments—they’re part of our daily routine. But starting a project often comes with a surprisingly stressful hurdle: finding the right teammates. Most team-building happens in messy Discord servers or Slack channels where you “post and pray,” hoping someone with the right skills, goals, and work style sees your message.

We didn’t just want a partner—we wanted a specific collaborator, like a “React developer passionate about Fintech” or an “ML engineer who already knows PyTorch.” When teams form randomly, mismatched expectations and roles can lead to confusion, missed deadlines, and burnout.

That’s why we built Pilot Episode: a structured, role-based matching platform that helps students find collaborators based on technical fit, interests, and project needs—so the “first episode” of every great student project starts with clarity, momentum, and healthier teamwork.

What it does

Pilot Episode is a specialized matching platform that helps students turn ideas into teams. Unlike general bulletin boards, it allows project leads to define specific roles (e.g., Frontend, Backend, ML Engineer) and set required headcounts.

Granular Discovery: Users can filter projects by 10 specialized categories and search for specific tools in the 33+ tech stack database.
Structured Applications: Students apply for specific roles within a project, and the system tracks their application status in real-time.
Personalized Dashboard: A centralized hub where users manage the projects they’ve led and track the progress of every application they've submitted.

How we built it

We engineered a high-performance stack that reflects modern industry standards:

Frontend: Developed using React (Vite) for a lightning-fast SPA experience. We utilized Custom Vanilla CSS instead of external libraries to maintain full control over the UI/UX and unique brand identity.
Backend & Auth: We implemented Supabase for PostgreSQL database management and user authentication.
Database & Storage: We designed a normalized relational schema to handle complex many-to-many relationships. Supabase Storage was integrated for real-time image hosting of user avatars and project assets.

Challenges we ran into

The most significant hurdle was the Relational Logic of Role-Based Applications. Managing a system where $Applications = Users \times Roles$ required complex foreign key constraints and careful state management. For instance, when a project lead closes a specific role, the system must instantly update the status for all pending applicants. Additionally, ensuring a smooth State Synchronization between the Supabase real-time database and our React frontend without unnecessary re-renders was a constant technical challenge.

Accomplishments that we're proud of

We are incredibly proud of building a 100% Custom UI/UX. By avoiding CSS frameworks like Tailwind or Bootstrap, we successfully built a premium, responsive interface from scratch. We also successfully implemented a robust Filtering & Search Algorithm that allows users to query not just by titles, but by individual technology tags nested within the project data.

What we learned

Relational Database Design: We gained deep experience in data normalization and maintaining referential integrity in a complex project ecosystem.
Advanced React Patterns: We mastered managing complex global states and handling asynchronous data fetching across multiple components.
Collaboration Workflow: Working as a team of four, we sharpened our skills in Git branch management and technical documentation to ensure a seamless integration of frontend and backend.

What's next for Pilot Episode

The journey doesn't end here. We plan to integrate the GitHub API to automatically verify a user's technical skills based on their repositories. We also aim to add an In-app Chat feature so that matched teammates can start brainstorming immediately. Finally, we want to expand the platform to support Course-Specific Matching to help students in large CS lectures find study groups and project partners more effectively.