What is Amazon SageMaker AI?

Current naming: AWS renamed Amazon SageMaker to Amazon SageMaker AI on December 3, 2024. The sagemaker API namespaces and existing feature names remain unchanged for backward compatibility.

• A fully managed machine-learning service for building, training, and deploying ML models.
• Supports ML workflows across development, training, evaluation, and production deployment.
• Supports managed algorithms as well as custom algorithms and frameworks.

Pricing of Amazon SageMaker

• For Free Tier Limits: https://aws.amazon.com/free
• For SageMaker Pricing: AWS SAGEMAKER PRICING

How to Use Amazon SageMaker

You have several options for how you can use Amazon SageMaker.

For Beginner Users

1. IDE: SageMaker Studio

• Web-based UI
• Seamless integration with deep learning and data science environments and scalable compute resources for training, inference, and other ML operations.
• No need for complex systems admin and security processes.
• Fully control data access and resource provisioning for users.

2. Notebook instances: SageMaker-managed ML compute instances with Jupyter environments.

• Notebook instances remain available, but for new notebook-based work AWS generally recommends the current SageMaker Studio experience.

For Advanced Users

3. Command line & SDK: AWS CLI, boto3, & SageMaker Python SDK

4. Third-party integrations: Kubeflow and Kubernetes integrations can be used when you need Kubernetes-based orchestration alongside SageMaker capabilities.