BuddyTutor

Team Photo

Inspiration

We all realized that we used chatGPT for assistance with coursework on a daily basis, so we figured that teachers would want to know how exactly we were using chatGPT to help us with our coursework. Additionally we figured that a tool that incorporates these results from the students would be useful for teachers to know how best to help their students and observe what topics they were struggling. This also benefits the students as this tool would be integrated into their course platform, being structured to prioritize guidance over direct answers, thus help them to actually learn and receive the same guidance that they would receive from a teacher or tutor.

What it does

We developed a web application that integrates chatGPT into a typical course platform, providing teachers with data and analytics from students' questions and answers. Also, it provided direct chatGPT integration into each assignment with prompt engineering that direct responses to students' questions to help guide them through each assignments or quiz.

How we built it

We used Next.js and tailwind-CSS for the front end, both web-development frameworks that are new to all members of the team. For the backend of our system, we used express.js and node.js for the chatGPT-API integration and Auth0 authentication. This is also connected to a MongoDB database, which is used to store course and assignment information, as well as the conversations between the students and ChatGPT.

Challenges we ran into

One major challenge was using NextJS with server side rendering, as it is new to all of us on the team so we struggled to develop with it but made it work. We also ended up converting from typescript to javascript with the front end. We also had networking issues with MongoDB and the backend REST API. And one final challenge was adapting the project to be deployed as there were many issues along the way with getting it deployed.

Accomplishments that we're proud of

Our favorite accomplishments:

Making a web app in a completely new framework with NextJS.
Integrating the backend with ChatGPT's API and curating the responses with prompt engineering
Working as a team in a very fast paced environment with many issues and roadblocks that we overcame together ## What we learned
It takes a lot of time and patience to learn something new
Most of our progress came near the end of the hackathon, even though we were exhausted
Working together is essential to overcoming challenges quickly and effectively
Don't plan on sleeping if you're learning a new framework during a hackathon
Full-stack development is challenging but extremely rewarding ## What's next for BuddyTutor In the future, we would like to move a highly scalable system that can handle many concurrent users at a time. Also, we want to improve the database and prompt engineering for data aggregation and analytics. And finally, we want to make this platform easy to integrate with other course platforms such as Canvas.