WorkWell AI - AI-Powered Workforce Well-being Dashboard

An intelligent healthcare analytics platform that combines burnout prediction, workload balancing, and real-time employee monitoring to optimize workforce well-being and operational efficiency.

🎯 Overview

WorkWell AI is a comprehensive workforce management system designed for healthcare facilities (hospitals, clinics) that:

• Predicts burnout risk using AI models trained on synthetic hospital data
• Automatically redistributes workload to prevent doctor/employee overload
• Monitors employee well-being in real-time using facial emotion recognition
• Provides actionable insights through interactive dashboards and visualizations

✨ Key Features

🤖 AI-Powered Burnout Prediction

• Trains lightweight AI models using 50+ synthetic doctor profiles
• Calculates burnout scores based on:
  • Weekly hours worked
  • Patient load
  • Emotional state (emotion scores)
• Categorizes risk levels: Low, Medium, High
• Formula: burnoutScore = (weeklyHours * 0.4) + (patientLoad * 0.4) + (emotionScore * 100 * 0.2)

⚖️ Smart Workload Balancing

• Analyzes real-time patient traffic and doctor availability
• Identifies overloaded doctors (burnoutScore > 70 or patientsToday > 40)
• Redistributes patients to less loaded doctors
• Prioritizes doctors with lower burnout scores
• Provides workload recommendations via API

🔄 Automatic Workload Scheduler

• Runs every 60 seconds automatically
• Redistributes workload based on burnout risk
• Updates database with optimized patient distribution
• Logs before/after states for transparency
• Manual trigger available via /api/ai/optimize

👥 Real-Time Employee Monitoring

• Multi-face detection using face-api.js
• Emotion recognition (happy, sad, angry, neutral, etc.)
• Burnout score calculation from facial expressions
• Live dashboard updates every 15 seconds
• Face recognition and employee tracking

📊 Interactive Dashboards

• Employee Overview: Card-based layout showing burnout scores, hours, and risk levels
• Workload Overview: Bar charts comparing current vs recommended patient loads
• Burnout Charts: Visual distribution of risk levels
• Real-time Metrics: Live updates on employee status

🛠️ Tech Stack

Frontend

• React 19.1.1 - UI framework
• Vite - Build tool and dev server
• Tailwind CSS 4.1.17 - Styling
• Recharts 3.3.0 - Data visualization
• Chart.js 4.5.1 - Chart components
• face-api.js 0.22.2 - Facial recognition and emotion detection

Backend

• Node.js - Runtime environment
• Express 5.1.0 - Web framework
• MongoDB - Database (with in-memory fallback)
• Mongoose 8.19.3 - ODM for MongoDB
• ES6 Modules - Modern JavaScript syntax

📦 Installation

Prerequisites

• Node.js 18+ installed
• MongoDB installed and running (optional - system works with in-memory fallback)

Step 1: Clone the Repository

git clone <repository-url>
cd workwell-ai

Step 2: Install Backend Dependencies

cd backend
npm install

Step 3: Install Frontend Dependencies

cd ../frontend
npm install

Step 4: Configure Environment Variables

Create backend/.env:

PORT=5000
MONGODB_URI=mongodb://localhost:27017/workwell-ai
NODE_ENV=development

Step 5: Start MongoDB (Optional)

# Windows
mongod --dbpath C:\data\db

# Linux/Mac
sudo systemctl start mongod

🚀 Running the Application

Start Backend Server

cd backend
npm run dev

Expected output:

🚀 Server running on port 5000
🤖 Initializing AI System...
📊 Training burnout prediction model...
✅ AI System initialized successfully!
🤖 Starting AI Scheduler (runs every 60 seconds)...

Start Frontend Development Server

cd frontend
npm run dev

Expected output:

VITE ready in XXX ms
➜ Local: http://localhost:5174/

Access the Application

• Frontend Dashboard: http://localhost:5174
• Backend API: http://localhost:5000
• MongoDB: mongodb://localhost:27017/workwell-ai

📡 API Endpoints

Employee Management

• GET /api/employees - Fetch all employees
• POST /api/employees/update - Create or update employee with emotion data

AI & Analytics

• POST /api/ai/predictBurnout - Predict burnout risk for doctor(s)
• GET /api/ai/balanceWorkload - Get workload balance recommendations
• POST /api/ai/optimize - Manually trigger workload optimization
• GET /api/ai/telemetry - Get AI telemetry data
• POST /api/ai/telemetry - Send AI telemetry data

Time Tracking

• GET /api/time/today - Get today's time tracking data

🧠 AI Features Explained

1. Burnout Prediction Model

Located in backend/ai/modelTrainer.js:

• Reads training data from backend/data/doctorTrainingData.json
• Uses weighted formula to calculate burnout scores (0-100 scale)
• Stores predictions in MongoDB doctorAnalytics collection
• Falls back to in-memory storage if MongoDB unavailable

2. Workload Balancer

Located in backend/ai/workloadBalancer.js:

• Generates mock patient traffic data
• Combines with burnout scores to identify overloaded doctors
• Calculates optimal patient distribution
• Returns recommendations for workload redistribution

3. AI Scheduler

Located in backend/aiScheduler.js:

• Automatically runs every 60 seconds
• Fetches all doctors and their current workloads
• Identifies overworked (currentPatients > recommendedPatients) and underworked doctors
• Redistributes patients prioritizing lower burnout scores
• Updates database with new distribution
• Logs detailed before/after states

📁 Project Structure

workwell-ai/
├── backend/
│   ├── ai/
│   │   ├── modelTrainer.js          # AI burnout prediction model
│   │   └── workloadBalancer.js      # Workload balancing logic
│   ├── aiScheduler.js               # Automatic workload scheduler
│   ├── config/
│   │   └── db.js                    # MongoDB connection
│   ├── controllers/
│   │   ├── aiController.js          # AI endpoints controller
│   │   ├── employeeController.js    # Employee CRUD
│   │   └── timeController.js        # Time tracking
│   ├── data/
│   │   ├── doctorTrainingData.json  # 50+ synthetic doctor profiles
│   │   ├── generateDoctorData.js    # Data generator script
│   │   └── mockData.json            # Mock employee data
│   ├── models/
│   │   └── Employee.js              # Employee MongoDB schema
│   ├── routes/
│   │   ├── ai.js                    # AI routes
│   │   ├── employees.js             # Employee routes
│   │   └── time.js                  # Time routes
│   └── server.js                    # Express server
│
├── frontend/
│   ├── src/
│   │   ├── components/
│   │   │   ├── Dashboard.jsx        # Main dashboard
│   │   │   ├── EmployeeOverview.jsx # Employee cards
│   │   │   ├── WorkloadOverview.jsx # Workload charts
│   │   │   ├── BurnoutChart.jsx     # Burnout distribution
│   │   │   └── ...
│   │   ├── pages/
│   │   │   └── MonitorCam.jsx       # Webcam monitoring
│   │   ├── lib/
│   │   │   └── api.js               # API utilities
│   │   └── ...
│   └── ...
│
└── README.md

🎨 Frontend Components

EmployeeOverview

• Location: frontend/src/components/EmployeeOverview.jsx
• Features:
  • Responsive card grid layout (1/2/3 columns)
  • Displays: Name, Weekly Hours, Today Hours, Burnout Score, Risk Status
  • Auto-refreshes every 15 seconds
  • Color-coded risk badges (Red/Blue/Green)

WorkloadOverview

• Location: frontend/src/components/WorkloadOverview.jsx
• Features:
  • Bar chart comparing current vs recommended patient loads
  • Highlights overloaded doctors in red
  • Summary statistics cards
  • Auto-refreshes every 30 seconds

Dashboard

• Location: frontend/src/components/Dashboard.jsx
• Features:
  • Metric cards (Total Employees, High Risk Count, etc.)
  • Burnout distribution chart
  • Workload overview integration
  • Employee table

🔧 Configuration

MongoDB Connection

The system gracefully handles MongoDB connection:

• If MongoDB is available: Uses persistent storage
• If MongoDB is unavailable: Falls back to in-memory storage
• Connection string: mongodb://localhost:27017/workwell-ai

AI Scheduler Settings

• Interval: 60 seconds (configurable in server.js)
• Manual Trigger: POST /api/ai/optimize

Data Refresh Rates

• Employee Overview: 15 seconds
• Workload Overview: 30 seconds
• AI Scheduler: 60 seconds

📊 Data Models

Employee Schema

{
  name: String (required),
  email: String (unique, required),
  role: String (default: 'Employee'),
  burnoutScore: Number (0-100),
  status: String (Low/Normal/Medium/High),
  weeklyHours: Number,
  todayHours: Number,
  currentPatients: Number,
  recommendedPatients: Number,
  faceDescriptor: [Number], // 128-dim face embedding
  lastEmotion: String,
  emotionHistory: [{
    emotion: String,
    score: Number,
    timestamp: Date
  }],
  lastUpdated: Date
}

DoctorAnalytics Schema

{
  doctorId: Number (unique, required),
  name: String,
  department: String,
  burnoutScore: Number (0-100),
  category: String (Low/Medium/High),
  weeklyHours: Number,
  patientLoad: Number,
  emotionScore: Number,
  currentPatients: Number,
  recommendedPatients: Number,
  shiftDuration: Number,
  lastUpdated: Date
}

🧪 Testing

Test Burnout Prediction

curl -X POST http://localhost:5000/api/ai/predictBurnout \
  -H "Content-Type: application/json" \
  -d '{
    "doctor": {
      "doctorId": 1,
      "weeklyHours": 60,
      "patientLoad": 50,
      "emotionScore": 0.3
    }
  }'

Test Workload Balance

curl http://localhost:5000/api/ai/balanceWorkload

Trigger Manual Optimization

curl -X POST http://localhost:5000/api/ai/optimize

🚨 Troubleshooting

MongoDB Connection Issues

• Problem: "MongoDB Connection Error"
• Solution: System automatically falls back to in-memory storage. Check MongoDB is running: mongod --version

AI Model Not Training

• Problem: "Training data not found"
• Solution: Run node backend/data/generateDoctorData.js to generate synthetic data

Frontend Not Loading

• Problem: CORS errors or connection refused
• Solution: Ensure backend is running on port 5000 and frontend on 5174. Check backend/server.js CORS configuration.

🤝 Contributing

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

📝 License

This project is licensed under the ISC License.

🙏 Acknowledgments

• face-api.js - Facial recognition and emotion detection
• Recharts - Beautiful chart visualizations
• Tailwind CSS - Utility-first CSS framework
• MongoDB - NoSQL database

📧 Support

For issues, questions, or contributions, please open an issue on the repository.

---

Built with ❤️ for better workforce well-being