Comet Navigator

Comet Navigator is a student's all-in-one planner that automatically schedules their day. It fills open gaps with smart tasks, so students stay productive without the hassle.

Comment

Inspiration

College students today are overwhelmed with different technologies. Especially at the University of Texas at Dallas, students are constantly jumping between platforms such as eLearning for assignments, Outlook for announcements, club websites for events, personal calendars for scheduling, and group chats for study sessions. And the result? Students struggle to keep track of announcements, discover clubs and events, find study partners, and managing their time effectively. Ultimately, students miss opportunities simply because the information is scattered.

What it does

Our solution, Comet Navigator, is a dynamic student scheduling platform that guides students through their entire academic life! It combines scheduling, discovery, and AI recommendations into a single platform. In addition to similar regular scheduling platforms that optimize student's day to day life, Comet Navigator takes the extra step to actively help students navigate their university experience.

How we built it

Our frontend was built using NextJS with TypeScript to enable fast development, improved server side rendering, and easy API integration with our backend services. We designed the UI in Figma and rapidly implemented components which allowed us to assemble polished interfaces. Key features include:

  • Interactive Dashboard
  • Dynamic Calendar System
  • Study Room Finder
  • Club Discovery Interface
  • Friend System and Buddy Matching
  • Study Session Queue
  • Onboarding Survey
  • End of day Feedback Swipe UI

Figma Video: https://youtu.be/EZedzL_i8mk?si=-qyjfsLQM_E_k_uX Figma Link: https://www.figma.com/design/FBCu8cquDAygSMGi3pVdkj/Kitchen--Copy-?node-id=52-1646&t=2tLFXNJTdSpMDUwk-1

Our backend is powered by FastAPI which allowed us to build high performance APIs quickly while using Pydantic for strong data validation. FastAPI was ideal because it integrates well with AI pipelines and supports asynchronous requests. (mention Nebula APIs)

For our database, we used MongoDB to store user data such as:

  • User profiles
  • Survey responses
  • Calendar events
  • Study sessions
  • Friend relationships
  • Club memberships

We used Mongoose to manage our data models and enforce structure within the NoSQL environment.

And for our AI & data Processing side, we integrated Gemini API for recommendations and used NumPy for schedule analysis. The AI analyzes user availability, productivity, commitments, interests, and location to recommend study blocks, suggest campus events, match students with study partners, and optimize daily schedules.

And most importantly, we integrated University Data Sources such as Nebula Rooms and Nebula Clubs (Nebula API). Users could find nearby open study rooms and automatically assign rooms when creating study sessions. Users are also given recommendations on student organizations based on interests, major, availability, and past involvement while filtering our clubs the user already belongs to.

Challenges we ran into

The scraper was really hard to implement due to small intricacies such as element only loading when the user scrolled down or the html structure of the webpage being confusing to navigate and needing to find specific class or id tags which required allot of trial and error. Another big issue was the recommendation system needing to be tuned properly so that the user could get proper class time suggestion and club suggestion.

Accomplishments that we're proud of

One of our biggest accomplishments was designing a sleek, modern and intuitive frontend that makes navigating the platform feel effortless. We focused heavily on creating a clean user interface that students will actually enjoy using rather than something that felt cluttered or overwhelming.

What's next for Comet Navigator

We hope to deepen integrations with platforms like Canvas LMS so that assignments, deadlines, and announcements can automatically populate a students calendar in real time. This would allow the system to dynamically adjust study blocks and schedules as new coursework appears.

Built With

Share this project:

Updates