Gravitas

Know what's hard. Do it first.

Comment

Inspiration

Managing multiple courses and assignments can be overwhelming for university students. Platforms like Quercus and Crowdmark provide data, but it’s scattered and lacks actionable insights. We wanted to create a single, intelligent dashboard that not only aggregates assignments but also estimates workload, difficulty, and effort using AI, helping students plan better and avoid last-minute stress.

What it does

Gravitas centralizes all your assignments from Quercus and Crowdmark into a single sleek dashboard, automatically filtering out assignments without due dates or those that have already passed. Using AI (Google Gemini), it provides concise summaries, estimated completion times, and clear indicators of difficulty. The due dates automatically are populated into the calendar and is available for export into all calendar apps. The all in one simple, easy-to-use interface.

How we built it

• Frontend: Next.js, React, and Tailwind CSS, Shadcn for a responsive, clean UI.
• Backend & Integration: Next.Js,  ViteReact for Chrome Extensions, Custom clients for Quercus and Crowdmark to fetch courses, assignments, and statistics.
• AI Analysis: Google Generative AI (Gemini) for assignment summaries and workload estimation. We implemented throttling to avoid overwhelming the AI API.
• Utilities: PDF and HTML text extraction to analyze assignment content effectively.

Challenges we ran into

• Handling assignments with missing or malformed due dates.
• Handling assignments with .pdf file instructions or embedded links
• Throttling AI requests to stay within API limits while maintaining responsiveness.
• Parsing and normalizing data from multiple sources (Quercus vs Crowdmark) with different formats.

Accomplishments that we’re proud of

• Successfully combined multiple course platforms into a single dashboard.
• Developed an AI-powered workload and difficulty estimator that helps students prioritize assignments.
• Built a system that’s robust to missing or inconsistent data and dynamically filters out irrelevant assignments.

⸻

What we learned

• The importance of data normalization when integrating multiple APIs.
• How to use AI responsibly by throttling requests and handling errors gracefully.
• Building a UX that is both sleek and functional improves engagement, even for complex information.

⸻

What’s next for Gravitas

• Real-time notifications for upcoming assignments and deadlines.
• Support for more platforms beyond Quercus and Crowdmark.
• Deeper AI analysis: breaking down assignments into subtasks and estimating effort per step.
• Manual instructional/description for assignments

Built With

Share this project:

Updates