CYRUS’SCloud CornerTECHNOLOGY · TRUST · PERSPECTIVE
Home / AI / AWS Cloud AI

AWS Cloud AI

Amazon Bedrock supports foundation-model applications; Amazon SageMaker AI supports custom model development and operations. Choose according to workload, data, and operating requirements.

AWS cloud AI illustrated with Data foundations, Amazon Bedrock, SageMaker AI, Managed AI APIs
AWS cloud AI — an illustrated overview of the topics on this page.

Content reviewed: October 25, 2026

Amazon Bedrock

Use Bedrock for supported model access and services around generative applications. Model availability, features, and deployment options differ by model and region. Evaluate the specific combination you intend to use rather than treating the catalog as uniform.

Amazon SageMaker AI

SageMaker AI provides tools for preparing, training, deploying, and operating machine-learning models. It serves a different engineering need from simply invoking a managed foundation model. Match training and inference choices to the team’s ability to maintain data, models, and endpoints.

Choose the AWS AI delivery path: Application task — Define a measurable outcome; Managed model access — Evaluate Bedrock configurations; Custom ML lifecycle — Assess SageMaker AI requirements; Release evidence — Check data, quality and operations
Choose the AWS AI delivery path · Conceptual diagram; see the worked example and references.

Secure the application

  • Use scoped IAM identities and protect the data supplied to models.
  • Separate retrieval permissions from generated tool suggestions.
  • Review retention, encryption, network access, and logging for each chosen service.
  • Test prompt injection, unsupported answers, and tool failures.
  • Track model version, latency, cost, and quality together.

Lifecycle and availability

Do not design new systems around SageMaker Edge Manager: it reached end of life in April 2024. Validate supported alternatives against your device, runtime, and management requirements. Services and individual models can have different retirement schedules.

Sources & further reading