Testing Professor
Testing ProfessorMulti-Cloud Guidebook · AWS
All services

Machine Learning

Amazon SageMaker

Build, train and deploy ML models at scale.

Official docs

Overview

SageMaker covers the full ML lifecycle — Studio IDE, training jobs, hyperparameter tuning, model registry and endpoints (real-time/serverless/batch).

When to use it

  • Train custom models
  • Host inference endpoints
  • MLOps with pipelines + registry

Setup

  1. Open SageMaker Studio (domain + user profile).
  2. Use built-in algorithms or bring your own container.
  3. Deploy to an endpoint with autoscaling.

How to use

Invoke endpoint
aws sagemaker-runtime invoke-endpoint --endpoint-name fraud --body fileb://payload.json --content-type application/json out.json

QA use cases

  • Shadow-deploy a new model variant and run QA inference comparisons before promoting.