SquadGame

Inspiration

We were inspired by the struggles of forming teams for hackathons, especially in remote settings like Zoom. Without an easy way to find teammates, many participants either scrambled to join last-minute teams or dropped out entirely — causing lower event engagement and financial loss for organizers. We wanted to solve this by making team formation simple, fast, and fun.

What we learned

Through user interviews and direct experience, we learned that people are often reluctant to commit right away. Instead, they prefer to explore different teammates and project ideas before deciding. We also realized that most participants finalize their team choices before the official deadlines, emphasizing the need for early, high-quality recommendations.

How we built it

We kicked off the project by brainstorming ideas and conducting interviews with participants and organizers to better understand the pain points. From there, we divided our work into four main streams:

Product Design: Designing a simple and intuitive matching experience.
Workflow Implementation: Building the AI matching engine and user interface.
User Interviews: Continuously gathering feedback to refine our approach.
Manual use case testing: Manually interact with Generative AI to validate the feasibility and usefulness of the product.
Pitch Deck Creation: Clearly communicating the value of our solution.

Challenges We Faced

One of the biggest technical challenges was building a high-accuracy matching algorithm. Rather than trying to manually tune complex models, we pivoted to a one-shot Generative AI approach, where all user profiles and project needs are fed into an LLM, which then generates tailored team recommendations.