Cal Poly Pomona Dining Dashboard

A sophisticated web-based dashboard for predicting staffing requirements at Cal Poly Pomona's dining facilities. Features real-time predictions, interactive visualizations, and an AI-powered assistant for intelligent insights.

🚀 Features

Core Functionality

• Real-time Predictions: Get staffing predictions based on weather, events, and dates
• Interactive Dashboard: Visual representation of staffing needs with stacked bar charts
• Quick Information: Today's summary with worker count, expected customers, and total hours
• Batch Predictions: Generate predictions for date ranges
• Detailed Analysis: Click on chart bars for in-depth breakdowns with weather and event scenarios

AI-Powered Assistant

• Intelligent Chat Interface: AI agent powered by AWS Bedrock and LangChain
• Contextual Responses: Understands staffing data, weather impacts, and event considerations
• Historical Data Access: Query past performance and trends
• Prediction Explanations: Get insights into why certain staffing levels are recommended
• Natural Language Queries: Ask questions in plain English about dining operations

Enhanced User Interface

• Modern Design: Cal Poly Pomona branded interface with smooth animations
• Responsive Layout: Works seamlessly on desktop, tablet, and mobile devices
• Interactive Charts: Click-to-explore functionality with hover details
• Accessibility Features: Keyboard navigation, focus indicators, and screen reader support

🤖 AI Assistant Capabilities

The integrated AI assistant can help with:

• Staffing Predictions: "What staffing do I need for next week if it's sunny with no events?"
• Weather Impact Analysis: "How does rainy weather affect dining traffic?"
• Event Planning: "How many extra workers do I need for graduation?"
• Operational Guidance: "How should I schedule my employees for optimal coverage?"

Chat Features

• Smart Fallbacks: Helpful responses even when AI agent is unavailable

📊 Worker Types Predicted

• General Purpose Worker (FOH_General)
• Cashier (FOH_Cashier)
• Chef (Kitchen_Prep)
• Line Workers (Kitchen_Line)
• Dishwasher (Dish_Room)
• Management (Management)

🌤️ Weather Conditions

• Sunny
• Cloudy
• Rainy
• Extreme Heat

🎓 Campus Events

• Regular Day
• Club Fair
• Career Fair
• Sports Events
• Graduation
• Parent Weekend
• Prospective Student Day
• Conference Hosting
• Campus Construction

🛠️ Setup Instructions

Prerequisites

• Python 3.8 or higher
• AWS Account with Bedrock access (for AI features)
• AWS CLI configured with appropriate credentials

1. Install Dependencies

pip install -r requirements.txt

2. AWS Configuration (Required for AI Assistant)

# Install AWS CLI if not already installed
pip install awscli

# Configure your AWS credentials
aws configure

You'll need:

• AWS Access Key ID
• AWS Secret Access Key
• Default region (e.g., us-west-2)
• Bedrock model access for Claude 3.5 Sonnet

3. Enable AWS Bedrock Access

1. Go to AWS Bedrock console
2. Navigate to "Model access" in the left sidebar
3. Request access to "Claude 3.5 Sonnet" model
4. Wait for approval (usually immediate for most accounts)

4. Verify Model Files

Ensure these files are present in your project directory:

• tx_model.pkl - Transaction prediction model
• work_model.pkl - Staffing prediction model
• df.csv - Historical data file
• inference.py - Prediction logic

5. Run the Application

python app.py

6. Access the Dashboard

Open your browser and go to http://localhost:5000

🔧 Configuration Options

Environment Variables

# AWS Configuration
export AWS_ACCESS_KEY_ID=your_access_key
export AWS_SECRET_ACCESS_KEY=your_secret_key
export AWS_DEFAULT_REGION=us-west-2

# Optional: Custom model paths
export TX_MODEL_PATH=./tx_model.pkl
export WORK_MODEL_PATH=./work_model.pkl
export DATA_FILE_PATH=./df.csv

Agent Configuration

The AI agent can be customized in app.py:

dining_agent = DiningHallAgent(
    enable_tracing=False,
    recursion_limit=50,  # Adjust for complex queries
    model_id="us.anthropic.claude-3-5-sonnet-20241022-v2:0"
)

🌐 API Endpoints

Dashboard APIs

• GET / - Main dashboard page
• GET /api/weather-options - Get available weather conditions
• GET /api/event-options - Get available campus events
• POST /api/predict - Single day prediction
• POST /api/batch-predict - Date range predictions
• GET /api/today-summary - Quick summary for today
• POST /api/simple-batch-predict - Simplified batch predictions
• POST /api/detailed-predict - Detailed single-day analysis

AI Chat APIs

• POST /api/chat - Send message to AI assistant
• POST /api/chat/clear - Clear conversation history
• POST /api/chat/reset - Reset agent state
• GET /api/chat/status - Check agent availability

💻 Usage Guide

Quick Start

1. View Today's Predictions: Automatically loaded on dashboard load
2. Generate Custom Predictions:
   • Select start and end dates
   • Click "Generate Predictions"
3. Explore Details: Click on any chart bar for detailed analysis
4. Chat with AI: Click the chat button (🤖) in bottom-right corner

Advanced Features

• Resize Chat: Drag the bottom-left corner of the chat window
• Full-Screen Chat: Press F11 or click the expand button (⛶)
• Keyboard Navigation: Use Tab, Enter, and Escape keys
• Mobile Support: Fully responsive design for all devices

AI Assistant Tips

• Ask specific questions about dates, weather, or events
• Request explanations for predictions
• Inquire about historical trends
• Get operational recommendations
• Use natural language - no special syntax required

📁 File Structure

├── app.py                    # Flask application with AI integration
├── inference.py              # Prediction logic
├── dining_agent.py           # AI agent implementation
├── cpp_agent.py              # Agent initialization
├── dataset_generator.py      # Data generation utilities
├── tx_model.pkl             # Transaction prediction model
├── work_model.pkl           # Staffing prediction model
├── df.csv                   # Historical data
├── requirements.txt         # Python dependencies
├── templates/
│   └── dashboard.html       # Enhanced dashboard template
├── static/
│   ├── css/
│   │   └── style.css        # Modern styling with Cal Poly colors
│   ├── js/
│   │   ├── dashboard.js     # Dashboard functionality
│   │   └── chat.js          # AI chat widget
│   └── images/
│       └── cpp_logo.svg     # Cal Poly Pomona logo
└── README.md               # This file

🎨 Design System

Color Scheme

The dashboard uses Cal Poly Pomona's official colors:

• Primary Green: #1B5E20
• Accent Green: #4CAF50
• Highlight Yellow: #FFC107
• Background: #F8F9FA
• Text: #495057

Typography

• Font Stack: -apple-system, BlinkMacSystemFont, 'Segoe UI', 'Roboto'
• Responsive Sizing: Scales appropriately across devices
• Accessibility: High contrast ratios and readable font sizes

🔧 Technologies Used

Backend

• Flask: Web framework
• Python: Core language
• AWS Bedrock: AI model hosting
• LangChain: AI agent framework
• LangGraph: Agent workflow management

Frontend

• HTML5: Modern markup
• CSS3: Advanced styling with Grid and Flexbox
• JavaScript: Interactive functionality
• Chart.js: Data visualization

Machine Learning

• XGBoost: Prediction models
• scikit-learn: Data processing
• pandas: Data manipulation
• numpy: Numerical computations

AI & NLP

• AWS Bedrock: Claude 3.5 Sonnet model
• LangChain: Agent framework
• boto3: AWS SDK

🚨 Troubleshooting

Common Issues

AI Agent Not Working

# Check AWS credentials
aws sts get-caller-identity

# Verify Bedrock access
aws bedrock list-foundation-models --region us-west-2

Recursion Limit Errors

• The system automatically handles these with fallback responses
• Clear chat history if issues persist
• Use /api/chat/reset endpoint to reset agent state

Model Files Missing

Ensure all .pkl files are in the project root directory

Port Already in Use

# Kill process on port 5000
lsof -ti:5000 | xargs kill -9

# Or use a different port
python app.py --port 5001

Performance Optimization

• Clear chat history periodically for better performance
• Use specific date ranges rather than very large ranges
• Monitor AWS Bedrock usage to manage costs

📈 Future Enhancements

• Multi-location Support: Extend to other dining facilities
• Advanced Analytics: Trend analysis and forecasting
• Mobile App: Native iOS/Android applications
• Integration APIs: Connect with existing campus systems
• Real-time Updates: Live data feeds and notifications

🤝 Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Test thoroughly
5. Submit a pull request

Built with ❤️ for Cal Poly Pomona Dining Services