🧠 ProfAIcademy

🚀 Inspiration

Education often lacks real-time personalization and insight, especially for students juggling multiple courses and deadlines. We wanted to build an AI-powered academic assistant that not only generates quizzes from study material but also tracks assignments, offers intelligent insights, and adapts to each student’s learning path.

💡 What it does

ProfAIcademy is an academic companion that helps students: • 📄 Upload PDFs or Images of notes and auto-generate multiple-choice quizzes. • 📆 Sync Courses and Assignments from Canvas LMS. • 💬 Chat with an AI Assistant that understands their workload and performance using OpenAI/Gemini. • 📊 Track Quiz Results and receive feedback on weak areas. • 🔐 Securely manage user credentials with JWT-based authentication.

🛠️ How we built it

We used a modern and modular tech stack: • Frontend: React (Next.js) + Tailwind CSS • Backend: FastAPI • Database: MongoDB Atlas • LLMs: OpenAI (gpt-3.5-turbo) + Gemini 1.5 Pro • Canvas LMS: Integrated via API

📦 MongoDB Collections

We heavily relied on MongoDB to structure and persist academic data:

Collection and Purpose for MongoDB Atlas DB Collections

users Stores hashed usernames & passwords

profile Contains user profile info from Canvas

courses Synced Canvas course data per user

assignments Synced assignments with due dates, grading type, etc.

quizzes Stores AI-generated quiz questions

chats
Optionally stores chat logs with AI

🧗 Challenges we ran into

• ⏳ Managing token expiration and syncing Canvas data securely
• 🧠 Parsing AI responses reliably into structured quizzes
• 🔐 Securely storing and validating user sessions with JWT
• 🌍 Deploying both backend and frontend while supporting real-time chat

🏆 Accomplishments that we’re proud of

• Successfully integrated LLMs with context-aware prompts
• Real-time assignment tracking and quiz generation
• Used MongoDB as a central backbone to power personalized learning
• Created a chatbot that gives meaningful academic feedback based on live data

📚 What we learned

• How to build a full-stack app with secure login, external API integrations, and LLMs
• Crafting system prompts for LLMs to generate structured and relevant output
• MongoDB indexing and schema flexibility for educational applications
• Parsing PDF and image data effectively to extract useful content

🔮 What’s next for ProfAIcademy

• 🎯 Add support for grading insights based on assignment scores
• 📱 Launch mobile version for students on-the-go
• 📊 Build a student dashboard with visual analytics and progress tracking
• 🤝 Allow teachers to input and validate quiz content
• 🧩 Integrate more LMS platforms like Blackboard or Moodle

Built With

Share this project:

Updates

posted an update

Another challenging part was the token quota on Gemini. I switched with the OpenAI for Chatbot since it was interacting with MongoDB Atlas data and that increased the token usage. Gemini was used for quiz generation.

Log in or sign up for Devpost to join the conversation.