AI Learning Planner 🚀

An AI-powered personalized learning platform that creates adaptive learning roadmaps using LLMs (Gemini / Ollama) based on user goals, experience, and available time.

---

✨ Features

• 🔐 Authentication (Register / Login / Logout)
• 🤖 AI-generated questionnaires (Gemini API / Local Ollama)
• 📚 Personalized weekly learning plans
• 🧭 Track progress (week completion)
• 👤 User profile management
• 🌍 Full-stack deployment on Vercel
• ⚙️ CI/CD with GitHub Actions
• 🧪 Automated backend testing (Jest)
• 🍪 Secure cookie-based auth (JWT)
• 🗃️ MongoDB persistence
• 🐳 Docker support for local setup
• 💻 Local AI support with Ollama (offline mode)

---

🧱 Tech Stack

Frontend

• React (Vite)
• Tailwind CSS
• Axios
• React Router
• Vercel

Backend

• Node.js
• Express
• MongoDB (Mongoose)
• JWT Authentication
• Google Gemini AI (cloud)
• Ollama (local AI)
• Serverless (Vercel)

DevOps

• GitHub Actions (CI/CD)
• Vercel Deployments
• MongoDB Atlas
• Docker (local development)

---

📁 Project Structure

AI-Learning-Planner/
├── client/
│   ├── public/
│   ├── src/
│   │   ├── api/            # Axios instance
│   │   ├── components/     # Reusable UI components
│   │   ├── pages/          # Route pages
│   │   │   ├── auth/       # Login/Register pages
│   │   │   ....            # Other Pages
│   │   │   .... 
│   │   ├── store/          # Redux tool kit
│   │   ├── App.jsx
│   │   └── main.jsx
│   ├── .env.development
│   ├── .env.production
│   └── vite.config.js
│
├── server/
│   ├── controllers/        # Business logic
│   ├── routes/             # API routes
│   ├── models/             # Mongoose models
│   ├── middleware/         # Auth, error handling
│   ├── config/             # CORS, env configs
│   ├── db/                 # MongoDB connection
│   ├── tests/              # Jest tests
│   ├── app.js              # Express app
│   └── server.local.js            # Server entry
│
├── docker-compose.yml
├── .github/
│   └── workflows/
│       └── ci-cd.yml
│
└── README.md

---

⚙️ Environment Variables

Backend (server)

.env.local

NODE_ENV=development
PORT=5000
MONGO_URI=mongodb://localhost:27017/ai_learning_local
JWT_SECRET=dev_secret
CLIENT_URL=http://localhost:5173
GOOGLE_API_KEY=your_key
REDIS_URL=XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
UPSTASH_REDIS_REST_URL=XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
UPSTASH_REDIS_REST_TOKEN="XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX"

---

.env.production (Vercel)

NODE_ENV=production
MONGO_URI=mongodb+srv://<username>:<password>@cluster.mongodb.net/ai_learning
JWT_SECRET=prod_secret
CLIENT_URL=https://ailearningplan.vercel.app
GOOGLE_API_KEY=your_key
REDIS_URL=XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
UPSTASH_REDIS_REST_URL=XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
UPSTASH_REDIS_REST_TOKEN="XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX"

---

Frontend (client)

.env.development

VITE_BACKEND_URL=http://localhost:5000

.env.production

VITE_BACKEND_URL=https://ai-learning-planner.vercel.app

---

🚀 Running Locally

1️⃣ Clone repository

git clone https://github.com/SARVESHYOGI/AI-Learning-Planner.git
cd ai-learning-planner

---

2️⃣ Start Backend

cd server
npm install
npm run dev

Runs on:

http://localhost:5000

---

3️⃣ Start Frontend

cd client
npm install
npm run dev

Runs on:

http://localhost:5173

---

🐳 Run with Docker (Optional)

docker compose up --build

---

🤖 Local AI Setup (Ollama)

If you want to use AI locally (offline, no cloud API, no cost), you can run the project with Ollama.

👉 Follow the full step-by-step guide here:
📄 Local AI Setup Guide

This mode is useful if:

• You want offline AI
• You don’t want to use Gemini API
• You want full local control
• You are developing without internet

⚠️ No environment variable change is required.
You only need to switch code in generator() function.

---

🧪 Running Tests

cd server
npm test

---

🔁 CI/CD Pipeline

Runs automatically on push to main.

Pipeline Steps:

1. Install dependencies
2. Run backend tests
3. Build frontend
4. Deploy to Vercel

---

🌍 Deployment URLs

Frontend

https://ailearningplan.vercel.app

Backend

https://ai-learning-planner.vercel.app

---

⏱️ Timeout Configuration

AI responses may take time (LLM generation).

Axios timeout:

timeout: 180000

(3 minutes)

---

🔐 Authentication Flow

• JWT stored in httpOnly cookies
• Secure cookies in production
• SameSite=None for cross-domain
• Axios uses withCredentials: true

---

🧠 AI Flow

1. User selects topic
2. AI generates questionnaire
3. User submits answers
4. AI generates learning plan
5. Plan stored in MongoDB
6. User tracks weekly progress

---

📌 Future Scope

• 💳 Payment integration (Stripe / Razorpay)
• 📱 Mobile app
• 📊 Analytics dashboard
• 🔔 Notifications
• 🧠 AI model switching UI
• 🗂️ Plan sharing
• 🏷️ Tags & goals

---

📜 License (IMPORTANT)

This project is licensed under the MIT License.

---

👨‍💻 Author

Sarvesh Yogi BE Computer Engineering, TSEC Full Stack Developer | AI & Backend Enthusiast

---

⭐ Support

If you find this project useful, please ⭐ the repository. PRs are welcome!