View the project demo here: https://hc-cdn.hel1.your-objectstorage.com/s/v3/9c0c68a2411c31f09a0ad439428b5e06e4f8f9c1_img_4771.mp4

Inspiration

At Achieve, staff members juggle spreadsheets and schedules to locate students. With so many students moving through classrooms on the Nobles campus, keeping track is inefficient and time-comnsuming, especially when its to urgently check in on a student.

What it does

Polaris is a student management system designed to help staff quickly locate students based on their daily schedules. It allows users to:

  • convert old spreadsheets of students to our system,
  • instantly find a student's current location based on their schedule.
  • view individual student schedules in a clean, easy-to-navigate format,
  • organize and classify student data into classrooms using AI-powered aggregation
  • access everything from a centralized dashboard, eliminating the need for scattered spreadshets.

How we built it

We started by generating structured example data using a large language model to simulate realistic student schedules. From there, we built a backend that classifies students into classrooms dynamically. Our frontend uses a clean and intuitive UI to make navigation seamless for staff. Technologies we used include Next.js and React for our website, and OpenAI to classify student data with ChatGPT.

Challenges we ran into

Determining a data structure for our student data that accurately reflects real-world constraints was an issue, and we wanted to ensure that student locations update efficiently without performance issues with a sleek UI.

Accomplishments that we're proud of

  • Successfully integrating AI to classify and organize student data
  • Building a fully-functional scheduling system in a short timeframe.
  • Designing an interface that balances simplicity and functionality

What we learned

  • How to use LLMs to generate and structure data for real-world applications.
  • Best practices for building user-friendly dashboards with React.
  • The importance of efficient data retrieval for real-time tracking applications.

Built With

Share this project:

Updates

Submission history