Inspiration

As students, we often lose focus during long classes, especially when learning in group settings where it can be challenging to ask for recaps or clarifications on what was just discussed. This gap in understanding can lead to missing key concepts and falling behind in the course material. We realized this issue is not unique to many students face similar difficulties, whether due to short attention spans or the complexity of the subjects being taught.

Recognizing this challenge, we wanted to explore how AI can enhance the learning experience, not just for students but also as a supportive tool for teachers. Rather than serving as a substitute for educators, we envisioned AI as a complement to traditional teaching methods in helping students stay engaged, fill in knowledge gaps, and ultimately deepen their understanding of the material.

What it does

Our project bridges the gap between eager students and expert tutors through a streamlined web application. Students can browse various topic categories and filter by expertise level to find the best tutor that matches their learning pace. Once a match is made, students and tutors can connect through video meetings, which are recorded for future reference. After each session, AI-generated notes provide concise summaries of the lessons, allowing students to review key points at any time. An accessible class history feature lets students revisit past sessions and notes, ensuring a continuous and personalized learning experience.

How we built it

Our project is built using a modern technology stack. For the frontend, we utilized Next.js with TypeScript to create a responsive and dynamic user interface. On the backend, we deployed a Linux server to host our REST API, developed with Flask, and connected it to a PostgreSQL database for efficient data management.

To enable video meetings between students and tutors, we integrated WebRTC, ensuring seamless real-time communication. For generating notes from these video sessions, we leveraged Google Cloud’s Speech-to-Text service to transcribe the conversations. These transcripts were then processed by the Gemini API, which created easy-to-read, concise notes, enhancing the review process for students.

Challenges we ran into

We faced many challenges while preparing and building this project. First of all we were unsure if the web-rtc sessions would work from different computers since we have never done any project with it before. Second we had some troubles installing packages and dependencies in the VPS, which he had to solve by changing the code to use different packages. Finally the biggest problem in a Hackathon is always the mental tiredness but we managed to overcome it.

Accomplishments that we're proud of

The biggest accomplishment we are proud of is being able to complete the project successfully. Because it is a very large project with different components we didn't thought we could complete. Then more specifically, we managed to make a successful rtc connection between different client with a live feed and audio. We also managed to connect it to a transcription tool and gemini to make a summary and key points for the classes.

What we learned

In this project we learned many different things, we created a web rtc and implemented it into a webpage, managed a server which in some moments we tought would break and everything would be over, but the thing we learnt the most was time micro management, we made sprints across this 24 hours in which we compared and helped each other out if we were struggling, teamwork was the key in delivering this project and we understood that since hour 1.

What's next for SabritO(n)es

We want to continue growing and challenging ourselves by entering different competitions like other Hackathons and really create applications and projects that will contribute to society in any way, we strive to create and learn new technologies so we can truly be ready for our future.

Share this project:

Updates