The problem PeerConnect.ai solves

Track: EduTech

Problem Statement

In educational settings, students often struggle to find suitable peers for collaborative learning and peer tutoring. Traditional methods of pairing students based on basic criteria like availability and subjects of interest often fall short in identifying the best matches that can truly enhance learning experiences. There is a need for a more sophisticated, AI-powered solution that can match students based on a deeper understanding of their strengths, weaknesses, learning styles, and feedback.

Problem It Solves

• Isolation: Addresses the issue of isolation among students by connecting them with potential learning partners.
• Finding Learning Partners: Simplifies the process of finding suitable study partners through detailed AI analysis.

Challenges we ran into

The biggest challenge we faced was that our team was formed just a day before the hackathon, so we had to quickly come up with an idea and ensure we could implement the core functionality of the web app. As college students with little to no experience in Next.js, we chose this stack to seize the opportunity to learn. The toughest part was integrating our backend logic with the Gemini AI model. We encountered difficulties in deciding on the appropriate prompt to achieve our functionality—finding the best peer match for a student among all listed students. We struggled to set up the correct prompt and eventually succeeded. The next hurdle was sending our student data to the model and letting it process it, which consumed much of our time. After achieving this, we faced the challenge of displaying the response to the user in an understandable way. The model's response was cluttered, necessitating a cleanup function that took considerable time to code, particularly the regex part. Despite completing the project in one day, we couldn't implement all the features we envisioned. Looking ahead, we aim to add comprehensive analytics to the UI using Chart.js and Ant Design UI, social platform links for students, real-time messaging with Socket.io, and video calling features. Our mentor also suggested onboarding users in a quiz format and gamifying the application with rewards for students. Our focus will be on continuing this project to make it a genuinely usable product for students.