awslabs/generative-ai-cdk-constructs

AWS Generative AI CDK Constructs are sample implementations of AWS CDK for common generative AI patterns.

What it solves

AWS Generative AI CDK Constructs provides a library of high-level, multi-service abstractions for the AWS Cloud Development Kit (CDK). It simplifies the process of building generative AI solutions by providing well-architected, repeatable infrastructure patterns in code, reducing the need for developers to manually configure complex AWS services for AI workloads.

How it works

The library extends the AWS CDK by offering "constructs"—pre-defined architectural patterns that bundle multiple AWS services with default configurations based on best practices. These are organized into L2 (lower-level) and L3 (higher-level) constructs:

  • L3 Constructs: Provide complete patterns for specific use cases, such as deploying foundation models from SageMaker JumpStart or Hugging Face, setting up Bedrock batch inference via Step Functions, or creating CloudWatch dashboards for Bedrock monitoring.
  • L2 Constructs: Provide simplified interfaces for specific AI-related AWS services, including Amazon Bedrock, Amazon OpenSearch Serverless Vector Collections, and Amazon S3 Vectors.

Who it’s for

Developers and cloud architects who use the AWS Cloud Development Kit (CDK) to build and deploy generative AI applications on AWS infrastructure.

Highlights

  • Multi-language support: Available for TypeScript, Python, Java, Go, and .NET (NuGet).
  • Pattern-based deployment: Includes ready-made patterns for SageMaker model deployment and Bedrock data automation.
  • Bedrock Integration: Specialized constructs for managing Amazon Bedrock and its associated vector databases (OpenSearch Serverless).
  • Well-architected: Default configurations are based on AWS best practices for predictable and repeatable infrastructure.