Amazon Bedrock Redundant API

This project implements a redundant API for Amazon Bedrock, allowing fallback to different models and regions.

Features

• Redundant model invocation across multiple AWS regions
• AWS X-Ray tracing integration
• REST API using Amazon API Gateway
• Serverless architecture using AWS Lambda

Prerequisites

• AWS SAM CLI installed
• Python 3.11 or later
• AWS credentials configured locally

Project Structure

.
├── config/
│   └── models.yaml          # Model configuration
├── src/
│   ├── lambda_function.py   # Lambda function code
│   └── requirements.txt     # Python dependencies (auto-installed by SAM)
├── tests/
│   └── test_lambda_function.py
└── template.yaml           # SAM template

Setup and Deployment

1. Install AWS SAM CLI:

   uv add aws-sam-cli

2. Configure your AWS credentials:

   aws configure

3. Build the application:

   sam build

4. Deploy the application:

   sam deploy --guided

   During the guided deployment, you will be prompted to:

   • Enter a unique stack name
   • Choose an AWS Region
   • Confirm creation of IAM roles

   The deployment will automatically:

   • Install Python dependencies from src/requirements.txt
   • Set up the API Gateway endpoint
   • Configure Lambda function with proper permissions
   • Set up X-Ray tracing
   • Load model configuration from config/models.yaml

Development

To run the tests locally:

python -m pytest tests/

To invoke the function locally:

sam local invoke BedrockRedundantFunction -e events/event.json

API Usage

Send a POST request to the deployed API endpoint with the following body structure:

curl -X POST \
  "<your api endpoint url>" \
  -H "Content-Type: application/json" \
  -d '{
    "contents": [
      "Tell me a joke"
    ],
    "temperature": 0.8,
    "max_tokens": 2048
  }'

The API will attempt to invoke the Bedrock models in order as defined in config/models.yaml, with automatic fallback if a model or region is unavailable.