Creek Monitoring Chatbot

This chatbot has been integrated into your Creek Monitoring Dashboard to provide intelligent assistance with water quality data analysis.

Features

🤖 Intelligent Data Analysis: Ask questions about water quality trends, site comparisons, and data interpretation

📊 Real-time Data Access: The chatbot has access to all your creek monitoring data including E. coli, pH, and turbidity measurements

🗺️ Interactive Dashboard Integration: Chatbot responses can trigger dashboard updates (e.g., selecting specific sites)

🔧 Multiple Tools:

• Site information and comparisons
• Water quality summaries
• Trend analysis
• Measurement explanations
• EPA standard compliance checking

Setup Instructions

1. Install Dependencies

The required packages have been added to requirements.txt. Install them with:

pip install -r requirements.txt

2. Set Environment Variables

You need to set the following environment variables:

export OPENAI_API_KEY="your-openai-api-key-here"

For deployment (e.g., Render), add this to your environment variables in the deployment settings.

3. Test the Chatbot

Run the test script to verify everything works:

python test_chatbot.py

4. Run the Dashboard

Start your dashboard as usual:

python app.py

Usage

In the Dashboard

1. Chat Interface: Located at the top of the dashboard

2. Quick Actions: Use the buttons for common queries:

   • "Water Quality Summary" - Get an overview of all sites
   • "Compare Sites" - Compare E. coli levels across sites
   • "Available Sites" - List all monitoring locations

3. Custom Questions: Type any question about:

   • Specific sites: "Tell me about Peavine creek/Old briarcliff way"
   • Trends: "Show me E. coli trends for the last 8 weeks at Lullwater creek"
   • Comparisons: "Which site has the highest turbidity?"
   • Standards: "Are any sites above EPA standards?"

Example Questions

• "What's the current water quality at Peavine creek/Oxford Rd NE?"
• "Compare pH levels across all sites"
• "Show me trends for turbidity at Lullwater creek"
• "Which sites are above EPA standards for E. coli?"
• "What does a pH of 7.2 mean for water quality?"
• "Give me a summary of the worst performing sites"

Technical Details

Architecture

• LangChain: Powers the conversational AI with OpenAI GPT-3.5-turbo
• Tools: Custom tools for data access and analysis
• Memory: Conversation history maintained for context
• Integration: Seamlessly integrated with Dash callbacks

Data Access

The chatbot has access to:

• All monitoring site data (E. coli, pH, turbidity)
• Site location information
• Historical trends and comparisons
• EPA standards and health implications

Customization

To add more functionality:

1. New Tools: Add methods to CreekDataTools class in chatbot.py
2. Enhanced Prompts: Modify the system prompt in CreekChatbot.__init__()
3. Additional Data: Extend the data processing in app.py

Troubleshooting

Common Issues

1. "OPENAI_API_KEY not set"

   • Ensure your environment variable is set correctly
   • Check deployment environment variables

2. "No data available"

   • Verify data files are in the correct location
   • Check file paths in the data loading section

3. Chatbot not responding

   • Check OpenAI API quota and billing
   • Verify internet connectivity
   • Check console for error messages

Debug Mode

Enable verbose logging by setting verbose=True in the AgentExecutor initialization in chatbot.py.

Future Enhancements

• Additional Knowledge Base: Integrate markdown/CSV files with interpretation guidelines
• Advanced Analytics: Add statistical analysis and forecasting
• Alert System: Notify users of concerning water quality changes
• Multi-language Support: Support for different languages
• Voice Interface: Add speech-to-text capabilities

Support

For issues or questions about the chatbot integration, check:

1. The console output for error messages
2. OpenAI API status and billing
3. Data file integrity and format