Inspiration
During our college experience, we have had problems easily finding events relevant to our interests as communication for some events are often sent to few people or get lost in the inbox. Building a schedule is also very stressful for many students as it is hard to predict how difficult the class will be for each unique student.
What it does
Helps students succeed by allowing them to build a user profile by adding their resume and current/past class schedule. An AI tool then processes this data and generates events by pulling data from current University of Michigan sources, and matches their profile to other students to give a glimpse into how difficult their classes may be.
How we built it
We used node.js for the backend which processed the json commands to the University of Michigan events. We then pulled class data for every class in the winter semester from the University of Michigan registrar. Our algorithm for classifying difficulty of classes used a weighted KNN algorithm to match similar student profiles. For the frontend, we used html to build the visuals and the different pages of the website.
Challenges we ran into
Correctly merging and sharing our files through Github. Parsing the resumes and storing data to be used in future algorithms in other parts of the website.
Accomplishments that we're proud of
The algorithm that would provide an idea of how difficult classes will be. Being able to store data on the server side which allows people to create and save a profile. Implementing apis to get json data and call AI agents.
What we learned
How to use html and javascript simultaneously. How to collaborate as a team and merge ideas at the same time without running into issues. How hosting a local server to create websites works.
What's next for MSuccess
Scaling to be able to work with the calendars and information from other colleges. Allowing students to implement their google calendar and schedules to allow for easier integration.
Log in or sign up for Devpost to join the conversation.