FairGrade

Promoting responsible AI use one student at a time

Comment

Inspiration

We saw a pervasive use of irresponsible AI use in education which hampers students abilities to properly learn. We wanted to find a way to promote responsible AI use among students while leveraging that data to do some good for educators as well. We wanted t achieve all this while creating an app that was simple and easy to use for educators.

What it does

FairGrade is multifaceted tool that allows students to leverage AI tools responsibly, without it impairing their learning. Additionally, using both chat data and grade data pulled from Moodle we generate reports regarding student strengths and weaknesses of either individual students or entire sections. This allows teachers to quickly tailor review sessions. Additionally teachers can create quizzes that can be taken directly on the app by students, with a chatbot available there to guide them and provide varying, if any, levels of help as decided by the teacher.

How we built it

We built the entire project using Next.js. Both the front-end pages and our back-end which connects to the necessary API are built using Next.js. For our database we chose to use Google's Firebase and Fire store. Our AI chatbot uses the Gemini API to generate messages. To run our own Moodle instance we used Linode and to connect to it, we used to Moodle API.

Challenges we ran into

We struggled a lot connecting to the Moodle API especially. The documentation proved to be lacking in some areas and it was difficult to connect. We also ran into many run of the mill issues, like forgetting awaits for our asynchronous statements.

Accomplishments that we're proud of

We were very proud to have achieved most of the goals we set out for the Hackathon. We were able to get the chatbot working properly, report generation for educators worked perfectly, and we were able to implement in most, but not all, of our objectives for in app quizzes.

What we learned

We learned that we need to be more familiar with the documentation of any external resources, such as APIs that we plan to use. At times we found ourselves confused on what was offered and what wasn't which cost us valuable time to figure out. In the future we would be more familiar with these before coming to a hackathon.

What's next for FairGrade

In the future we would like to have it so that quiz grades are sent directly to Moodle. Additionally we would like to implement even more connectivity with Moodle such as viewing grades in the application. We also wanted to implement a feature where the AI chatbots use only information found in the teacher's posted materials. This would prevent the chatbot from potentially using concepts or methods the student hasn't yet been exposed to.

Built With

Share this project:

Updates