Skip to main content

Amazon SageMaker

What is it​

A fully managed machine learning (ML) service that helps developers and data scientists build, train, and deploy ML models quickly.

What is it for​

Simplify the entire machine learning lifecycle, from data preparation to model deployment and monitoring in production.

Use cases​

  • Building and training ML models for various applications (computer vision, natural language processing, forecasting)
  • Deploying ML models in production for real-time or batch inference
  • Data preparation and engineering for ML
  • ML model experimentation and optimization
  • Building no-code ML solutions (with SageMaker Canvas)

Key points​

  • Fully managed: AWS handles the underlying infrastructure for model training and deployment
  • Comprehensive tools: Offers notebooks, pre-built algorithms, training and deployment environments, and monitoring tools
  • Scalability: Automatically scales computing resources for training and inference
  • Integration: Integrates with other AWS services like S3, Lambda, Glue, and ECR
  • SageMaker Studio: A unified IDE for the entire ML workflow
  • SageMaker Canvas: Allows business users to build ML models without writing code

Comparison​

  • Amazon SageMaker: Simplifies and accelerates ML development and deployment, abstracting infrastructure complexity. Ideal for teams that want to focus on data science and models.
  • ML training on EC2 (self-managed): Offers complete control over the environment but requires users to manually configure and manage servers, ML libraries, and tools. Can be more complex and time-consuming to set up and maintain.