Claude Search Assistant

An intelligent chatbot application that combines the power of Claude 3 Haiku with real-time web search capabilities, built with FastAPI and React.

Overview

Claude Search Assistant is a web-based chatbot that enhances Claude's capabilities with real-time web search functionality using the SERP API. This integration allows the chatbot to provide up-to-date information while maintaining Claude's conversational abilities.

Features

• Real-time Web Search: Integrates SERP API to fetch current information from the internet
• Intelligent Response System: Uses Claude 3.5 Haiku to determine when to use web search vs. model knowledge
• Simple Web Interface: Clean and responsive React-based chat interface
• Session Management: Maintains conversation history within sessions
• Rate Limiting: Built-in protection against excessive API usage
• Real-time Date Display: Shows current date for context awareness

Technology Stack

Backend

• FastAPI (Python web framework)
• LangChain (For tool integration and chat management)
• Anthropic's Claude 3 Haiku
• SERP API (For web search capabilities)
• AWS Bedrock (For Claude API access)

Frontend

• React
• Modern CSS with flexbox
• Real-time updates

Setup Instructions

Prerequisites

• Python 3.8+
• Node.js and npm
• AWS account with Bedrock access
• SERP API key
• Anthropic API key

Backend Setup

1. Navigate to the backend directory:

cd backend/app

2. Create a virtual environment:

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies:

pip install -r requirements.txt

4. Create a .env file in the backend/app directory with the following variables:

AWS_ACCESS_KEY_ID=your_aws_access_key
AWS_SECRET_ACCESS_KEY=your_aws_secret_key
AWS_DEFAULT_REGION=your_aws_region
SERP_API_KEY=your_serp_api_key

5. Start the backend server:

uvicorn main:app --reload

Frontend Setup

1. Navigate to the frontend directory:

cd frontend

2. Install dependencies:

npm install

3. Create a .env file in the frontend directory:

REACT_APP_API_URL=http://localhost:8000

4. Start the development server:

npm start

Usage Examples

1. Basic Questions

   • The chatbot will use Claude's knowledge base for general questions
   • Example: "What is the capital of France?"

2. Current Events

   • The system will automatically use web search for time-sensitive queries
   • Example: "What are the latest developments in AI technology?"

3. Real-time Information

   • Perfect for getting up-to-date information
   • Example: "What's the weather like in New York right now?"

4. Combined Knowledge

   • The system can blend Claude's knowledge with current web data
   • Example: "Compare historical and current inflation rates"

API Rate Limits

• The API is limited to 5 requests per minute per IP address
• Implement appropriate error handling in your applications

Contributing

1. Fork the repository
2. Create your feature branch: git checkout -b feature/YourFeature
3. Commit your changes: git commit -m 'Add YourFeature'
4. Push to the branch: git push origin feature/YourFeature
5. Submit a pull request

License

This project is open-source and available under the MIT License.

Future Improvements

• Add support for multiple search engines
• Implement user authentication
• Add support for file uploads and processing
• Enhance error handling and retry mechanisms
• Add support for streaming responses
• Implement conversation summarization
• Add support for multiple language inputs

Troubleshooting

Common issues and solutions:

1. API Connection Issues

   • Verify your environment variables are set correctly
   • Check your AWS credentials and permissions
   • Ensure SERP API key is valid

2. Rate Limiting

   • Implement exponential backoff in your requests
   • Consider caching frequent queries

3. Web Search Not Working

   • Verify SERP API key and quota
   • Check network connectivity
   • Review API response logs

For additional support, please open an issue in the GitHub repository.