Deployed link - https://study-buddy-nine-phi.vercel.app/
A full-stack web app that generates personalized learning paths, lets you chat with an AI tutor, and tests your knowledge with auto-generated quizzes — all powered by Groq's LLaMA model.
Built this as a learning project to get hands-on with the MERN stack. Ended up being one of the more interesting things I've worked on.
You type in a topic — say, "Machine Learning" or "System Design" — and the app generates a structured learning path broken into modules. From there you can:
- Chat with an AI tutor that knows your topic and remembers the conversation
- Mark modules as complete and track your progress
- Take a 5-question quiz generated specifically for your topic
- See detailed results with explanations for each answer
There's also an admin panel with basic analytics — total users, popular topics, average quiz scores, etc.
Backend
- Node.js + Express.js
- MongoDB + Mongoose
- JWT authentication
- Role-based access control (student / admin)
- Groq API (LLaMA 3.3 70B) for AI features
Frontend
- React + Vite
- React Router
- Axios
- react-hot-toast
- Pure CSS with CSS variables (light/dark mode)
- AI learning path generation — structured modules with estimated time
- AI tutor chat — full conversation history passed on every request
- AI quiz generation — MCQs with auto scoring and explanations
- Progress tracking — mark modules done, see % completion
- JWT auth — register, login, protected routes
- Dark mode — persisted to localStorage
- Admin dashboard — users, paths, analytics, popular topics
- Node.js v18+
- MongoDB (local or Atlas)
- Groq API key — free at console.groq.com
git clone https://github.com/yourusername/studybuddy-backend.git
cd studybuddy-backend
npm installCreate a .env file:
PORT=5000
MONGO_URI=mongodb://127.0.0.1:27017/StudyBuddy
JWT_SECRET=your_jwt_secret_here
JWT_EXPIRES_IN=7d
GROQ_API_KEY=your_groq_api_key_here
NODE_ENV=developmentnpm run devServer runs on http://localhost:5000
git clone https://github.com/yourusername/studybuddy-frontend.git
cd studybuddy-frontend
npm install
npm run devApp runs on http://localhost:5173
POST /api/auth/register
POST /api/auth/login
POST /api/paths/generate
GET /api/paths/my-paths
GET /api/paths/:id
PATCH /api/paths/:id/progress
POST /api/chat/:pathId
GET /api/chat/:pathId/history
POST /api/quiz/:pathId/generate
POST /api/quiz/:pathId/submit
GET /api/quiz/:pathId/result
GET /api/admin/users
GET /api/admin/analytics
DELETE /api/admin/users/:userId
studybuddy-backend/
├── config/ # DB connection
├── controllers/ # Route handlers
├── middleware/ # Auth + role middleware
├── models/ # Mongoose schemas
├── routes/ # Express routers
├── services/ # Groq AI integration
└── utils/ # JWT helper
studybuddy-frontend/
├── src/
│ ├── api/ # Axios instance
│ ├── components/ # Navbar, ThemeToggle, route guards
│ ├── context/ # Auth + Theme context
│ └── pages/ # All page components
- How conversation history works in LLM APIs — you pass the full message array on every request, there's no magic memory on the server side
- Why middleware order matters in Express — auth before role check, always
Promise.all()for parallel DB queries instead of sequential awaits- The difference between storing roles in JWT vs querying per request (performance vs freshness tradeoff)
MIT