Streamlined Amazon Bedrock Chatbot

A minimal Python-based chatbot implementation that supports both Amazon Nova and Anthropic Claude models through Amazon Bedrock.

Features

• Supports multiple foundation models:
  • Amazon Nova models (Micro, Lite, Pro)
  • Anthropic Claude models (3.7 Sonnet, 3.5 Haiku)
• Simple command-line interface
• Maintains conversation history
• Automatic AWS credential detection

Prerequisites

1. An AWS account with access to Amazon Bedrock
2. Model access enabled for Nova and Claude models
3. Python 3.9+
4. Boto3 library

Installation

1. Clone the repository

   git clone https://github.com/yourusername/bedrock-chatbot.git
   cd bedrock-chatbot

2. Create and activate a virtual environment

   # Create virtual environment
   python -m venv venv
   
   # Activate on macOS/Linux
   source venv/bin/activate
   
   # Activate on Windows
   venv\Scripts\activate

3. Install dependencies

   pip install -r requirements.txt

AWS Setup

Request Model Access

Before using the chatbot, request access to the models:

1. Sign in to the AWS Management Console
2. Navigate to Amazon Bedrock
3. Select "Model access" in the left navigation
4. Request access to:
   • Amazon Nova models (Micro, Lite, Pro)
   • Anthropic Claude models (3.7 Sonnet, 3.5 Haiku)

AWS Credentials

The chatbot uses the AWS SDK's default credential provider chain:

1. Environment variables
2. Shared credential file (~/.aws/credentials)
3. AWS config file (~/.aws/config)
4. IAM role for EC2/ECS

Usage

Make the script executable and run it:

chmod +x bedrock-chatbot.py
./bedrock-chatbot.py

Follow the prompts to:

1. Select a model
2. Type your messages
3. Type 'quit' to exit or 'clear' to start a new conversation

Available Models

Amazon Nova Models

• Nova Micro: Text-only model with lowest latency
• Nova Lite: Balanced multimodal model (images, text)
• Nova Pro: Most capable multimodal model

Anthropic Claude Models

• Claude 3.7 Sonnet: Advanced reasoning with standard and extended thinking modes
• Claude 3.5 Haiku: Fast, efficient model for everyday tasks

Troubleshooting

If you encounter errors:

1. Authentication errors: Verify AWS credentials are correctly configured
2. Access errors: Confirm model access has been granted in the Bedrock console
3. Region issues: The chatbot defaults to us-east-1; ensure your chosen models are available in this region

License

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