AccessiNotes Axel Rodriguez-Leon & Juan Saquino

Inspiration

We have both been TAs for a class that had a student who had requested a notetaker and we asked ourselves, what if there are no notetakers? What would happen if the class is too small or if the notetaker may ever need to be absent from a class? We wouldn't want them to be forced to be behind because of this.

What it does

This is a website that would connect to an AI that will be able to take in text files containing a transcript of the lecture and create a new set of notes. Additionally, the user can request to use the style of a notetaker in which the AI will analyze that style and generate a new set of notes using both the transcript and the notetaker's notes.

How we built it

We built the frontend through a combination of HTML, CSS, and Typescript. The backend was built upon the Google Gemma model which was run on a local machine with the help of Ollama. The model was able to connect to the local machine through a python script.

Challenges we ran into

We ran into some CSS issues in terms of setting up the website. Additionally, we ran into the issue of having to learn how to use an unknown AI API that had little documentation.

Accomplishments that we're proud of

Being able to have the AI model run on the local machine, especially for doing it for the first time and having to go through with little documentation.

What we learned

We learned how to connect a YouTube video through React/CSS as well as being able to set up an AI model on a local machine.

What's next for AccessiNotes

We would like to connect the frontend with the backend as well as maybe work with DSS to help improve upon the existing notetaker program.

Built With

Share this project:

Updates