AI-Powered Invoice Management System

A full-stack web application that uses AI to automatically extract, process, and manage invoice data using OCR and machine learning technologies.

🚀 Features

• AI-Powered OCR: DeepSeek OCR2 integration for accurate text extraction
• Intelligent Data Correction: LLM-based correction and structuring of OCR output
• Multi-Language Support: Automatic language detection and processing
• User Management: Role-based authentication (User/Admin)
• Real-time Analytics: Comprehensive dashboards with charts and insights
• Email Notifications: Automatic email alerts for processed invoices
• Camera Integration: Capture invoices directly from camera
• Editable Interface: Modify extracted data before saving
• Accuracy Tracking: Monitor OCR accuracy and processing performance

🛠 Tech Stack

Backend

• Node.js with Express.js
• MongoDB Atlas with Mongoose ODM
• JWT Authentication
• Multer for file uploads
• Nodemailer for email notifications
• Python Integration via child_process

Frontend

• React.js with React Router
• Tailwind CSS for styling
• Recharts for data visualization
• Lucide React for icons
• Axios for API calls
• React Hot Toast for notifications

AI/ML

• DeepSeek OCR2 for text extraction
• DeepSeek LLM for data correction
• Python scripts for AI processing

📁 Project Structure

invoice/
├── backend/
│   ├── src/
│   │   ├── controllers/     # API controllers
│   │   ├── middleware/      # Auth middleware
│   │   ├── models/         # Mongoose schemas
│   │   ├── routes/         # API routes
│   │   └── utils/          # Utility functions
│   ├── uploads/            # File upload directory
│   ├── .env                # Environment variables
│   └── package.json
├── frontend/
│   ├── src/
│   │   ├── components/     # React components
│   │   ├── contexts/       # React contexts
│   │   ├── pages/          # Page components
│   │   ├── services/       # API services
│   │   └── utils/          # Utility functions
│   ├── public/
│   ├── .env                # Environment variables
│   └── package.json
├── python/
│   ├── deep.py             # OCR processing script
│   └── llm_correct.py      # LLM correction script
└── README.md

🚀 Getting Started

Prerequisites

• Node.js (v14 or higher)
• Python (v3.7 or higher)
• MongoDB Atlas account
• Conda environment (recommended)

Installation

1. Clone the repository

   git clone <repository-url>
   cd invoice

2. Set up Python Environment

   conda create -n ocr python=3.8
   conda activate ocr

3. Install Python Dependencies

   cd python
   pip install -r requirements.txt

4. Install Backend Dependencies

   cd ../backend
   npm install

5. Install Frontend Dependencies

   cd ../frontend
   npm install

6. Environment Setup

   Backend (.env):

   # MongoDB Configuration
   MONGODB_URI=mongodb+srv://nithyashriskcs24_db_user:<password>@cluster0.tg9qtt0.mongodb.net/invoice_management
   
   # JWT Configuration
   JWT_SECRET=your-super-secret-jwt-key
   JWT_EXPIRE=7d
   
   # Server Configuration
   PORT=5000
   NODE_ENV=development
   
   # Email Configuration
   EMAIL_HOST=smtp.gmail.com
   EMAIL_PORT=587
   EMAIL_USER=your-email@gmail.com
   EMAIL_PASS=your-app-password
   
   # Python Scripts Path
   PYTHON_PATH=python
   DEEP_SCRIPT_PATH=../python/deep.py
   LLM_CORRECT_SCRIPT_PATH=../python/llm_correct.py

   Frontend (.env):

   REACT_APP_API_URL=http://localhost:5000/api
   REACT_APP_ENV=development

Running the Application

1. Start the Backend Server

   cd backend
   npm run dev

2. Start the Frontend Development Server

   cd frontend
   npm start

3. Access the Application

   • Frontend: http://localhost:3000
   • Backend API: http://localhost:5000
   • API Health Check: http://localhost:5000/api/health

📊 API Endpoints

Authentication

• POST /api/auth/signup - User registration
• POST /api/auth/login - User login
• GET /api/auth/profile - Get user profile
• PUT /api/auth/profile - Update user profile
• PUT /api/auth/change-password - Change password

Invoice Management

• POST /api/invoice/upload - Upload and process invoice
• PUT /api/invoice/:id/save - Save edited invoice
• GET /api/invoice/my - Get user's invoices
• GET /api/invoice/:id - Get specific invoice
• DELETE /api/invoice/:id - Delete invoice
• GET /api/invoice/analytics/my - Get user analytics

Admin APIs

• GET /api/admin/users - Get all users
• GET /api/admin/invoices - Get all invoices
• GET /api/admin/analytics - Get system analytics
• PUT /api/admin/users/:userId/status - Update user status
• PUT /api/admin/users/:userId/role - Update user role
• GET /api/admin/health - Get system health

🤖 AI Processing Pipeline

1. Image Upload: User uploads invoice image or captures via camera
2. OCR Processing: deep.py extracts raw text using DeepSeek OCR2
3. Data Correction: llm_correct.py structures and corrects OCR output
4. Accuracy Calculation: Compare original vs corrected data
5. Data Storage: Save structured data with metadata to MongoDB
6. Email Notification: Send processing summary to user

🔧 Configuration

Python Scripts

deep.py:

• Input: Image file path
• Output: Raw OCR text with metadata
• Integration: DeepSeek OCR2 API

llm_correct.py:

• Input: Raw OCR text
• Output: Structured JSON data
• Features: Field extraction, validation, language detection

Database Schema

User Model:

{
  username: String,
  email: String,
  password: String,
  role: String, // 'user' | 'admin'
  profile: {
    firstName: String,
    lastName: String,
    phone: String,
    company: String
  },
  preferences: {
    language: String,
    emailNotifications: Boolean,
    theme: String
  }
}

Invoice Model:

{
  userId: ObjectId,
  originalOCR: String,
  correctedData: Object,
  accuracyScore: Number,
  languageDetected: String,
  fileName: String,
  filePath: String,
  processingTime: Number,
  status: String,
  extractedFields: Object,
  metadata: Object
}

📈 Analytics Features

User Dashboard

• Total invoices processed
• Average OCR accuracy
• Processing time metrics
• Vendor insights
• Spending analytics

Admin Dashboard

• System-wide metrics
• User management
• Invoice volume trends
• Accuracy distribution
• Language distribution
• System health monitoring

🔐 Security Features

• JWT-based authentication
• Role-based access control
• Input validation and sanitization
• File upload security
• Rate limiting
• CORS protection
• Helmet.js security headers

📧 Email Integration

• Automatic notifications on invoice processing
• Processing summaries with accuracy metrics
• Configurable email templates
• Support for Gmail SMTP

🎨 UI/UX Features

• Responsive design for all devices
• Modern, clean interface with Tailwind CSS
• Interactive charts and data visualizations
• Real-time form validation
• Loading states and error handling
• Toast notifications
• Camera integration for mobile devices

🚀 Deployment

Production Setup

1. Environment Variables:

   • Set NODE_ENV=production
   • Update MongoDB URI with production database
   • Configure production email settings
   • Set strong JWT secret

2. Build Frontend:

   cd frontend
   npm run build

3. Deploy Backend:

   • Use PM2 for process management
   • Configure reverse proxy (Nginx)
   • Set up SSL certificates
   • Configure firewall rules

Docker Deployment (Optional)

# Backend Dockerfile
FROM node:16-alpine
WORKDIR /app
COPY package*.json ./
RUN npm ci --only=production
COPY . .
EXPOSE 5000
CMD ["npm", "start"]

🤝 Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Add tests if applicable
5. Submit a pull request

📝 License

This project is licensed under the MIT License - see the LICENSE file for details.

🆘 Troubleshooting

Common Issues

1. Python Scripts Not Found:

   • Ensure conda environment is activated: conda activate ocr
   • Check Python path in backend .env file

2. MongoDB Connection Error:

   • Verify MongoDB URI and credentials
   • Check network connectivity
   • Ensure IP is whitelisted in MongoDB Atlas

3. Email Not Sending:

   • Verify Gmail app password
   • Check SMTP settings
   • Ensure less secure apps are enabled

4. OCR Processing Fails:

   • Check Python script permissions
   • Verify image file format
   • Check DeepSeek API credentials

Debug Mode

Enable debug logging by setting:

NODE_ENV=development
DEBUG=invoice:*

📞 Support

For support and questions:

• Create an issue in the repository
• Check the troubleshooting section
• Review the API documentation

---

Built with ❤️ using Node.js, React, and AI technologies