Amazon SageMaker helps data scientists and developers to prepare, build, train, and deploy high-quality machine learning (ML) models quickly by bringing together a broad set of capabilities purpose-built for ML.
The most comprehensive ML service
Accelerate innovation with purpose-built tools for every step of ML development, including labeling, data preparation, feature engineering, statistical bias detection, auto-ML, training, tuning, hosting, explainability, monitoring, and workflows.
The first integrated development environment (IDE) for ML
Boost your productivity using Amazon SageMaker Studio, the first fully integrated development environment designed specifically for ML that brings everything you need for ML under one unified, visual user interface.
Functionality designed from the ground up to work together
Use Amazon SageMaker’s integrated capabilities for ML development, so you can eliminate months of writing custom integration code, and ultimately reduce cost.
How it works
-
Overview
-
Details
-
Overview
-
-
Details
-
SageMaker Capability Description Automatic Model Tuning Hyperparameter optimization Built-in and Bring-your-own Algorithms Dozens of optimized algorithms or bring your own Distributed training libraries - NEW
Training for large datasets and models Kubernetes & Kubeflow Integration Simplify Kubernetes-based machine learning Local Mode Test and prototype on your local machine Managed Spot Training Reduce training cost by 90% Multi-Model Endpoints Reduce cost by hosting multiple models per instance One-click Deployment Fully managed, ultra low latency, high throughput One-click Training Distributed infrastructure management SageMaker Autopilot Automatically create machine learning models with full visibility SageMaker Clarify - NEW
Detect bias and understand model predictions SageMaker Data Wrangler - NEW Aggregate and prepare data for machine learning SageMaker Debugger Debug and profile training runs SageMaker Edge Manager - NEW Manage and monitor models on edge devices
SageMaker Experiments Capture, organize, and compare every step SageMaker Feature Store - NEW Store, update, retrieve, and share features SageMaker Ground Truth Label training data for machine learning SageMaker JumpStart - NEW Pre-built solutions for common use cases SageMaker Model Monitor Maintain accuracy of deployed models SageMaker Pipelines - NEW Workflow orchestration and automation SageMaker Processing Built-in Python, BYO R/Spark SageMaker Studio Integrated development environment (IDE) for ML SageMaker Studio Notebooks Jupyter notebooks with elastic compute and sharing
One of the fastest growing services in AWS history
Amazon SageMaker is built on Amazon’s two decades of experience developing real-world machine learning applications, including product recommendations, personalization, intelligent shopping, robotics, and voice-assisted devices.
10x
increase in team productivity
90%
cost reduction with managed spot training
75%
lower inference costs
54%
70%
198
22
Amazon SageMaker supports the leading machine learning frameworks and toolkits
Key features to prepare data, and build, train, and deploy ML models
Improve productivity using the first fully integrated development environment (IDE) for ML
Amazon SageMaker Studio provides a single, web-based visual interface where you can perform all ML development steps required to prepare data, and build, train, and deploy models.
Build, train, and tune models automatically
Amazon SageMaker Autopilot selects the best algorithm for the prediction, and automatically builds, trains, and tunes machine learning models without any loss of visibility or control.
Reduce data labeling costs by up to 70%
Amazon SageMaker Ground Truth makes it easy to more accurately label training datasets for a variety of use cases including 3D point clouds, video, images, and text.
The fastest and easiest way to prepare data for ML
Amazon SageMaker Data Wrangler reduces the time it takes to prepare data for ML from weeks to minutes. With a few clicks, you can complete each step of the data preparation workflow, including data selection, cleansing, exploration, and visualization.
Purpose-built feature store for ML
Amazon SageMaker Feature Store provides a repository to store, update, retrieve, and share ML features. SageMaker Feature Store offers one consistent view of features for ML models to use so it becomes significantly easier to generate models that produce highly accurate predictions.
Train high-quality models faster
Amazon SageMaker provides a built-in debugger and a profiler so you can identify and reduce training errors and performance bottlenecks from your models before pushing them to production.
One-click deployment to the cloud
Amazon SageMaker makes it easy to deploy your trained model to production with a single click, so you can start generating predictions for real-time or batch data.
Improve quality of models on edge devices
Amazon SageMaker Edge Manager helps you optimize, secure, monitor, and maintain machine learning models on fleets of edge devices to ensure that models deployed on edge devices are operating correctly.
Essential features for ML in production
Automate machine learning workflows
Amazon SageMaker Pipelines is the first purpose-built, easy-to-use continuous integration and continuous delivery (CI/CD) service for machine learning. Workflows can be shared and re-used between teams.
Detect bias and understand predictions
Amazon SageMaker Clarify provides bias detection across the ML workflow, enabling you to build greater fairness and transparency into your ML model. SageMaker Clarify also includes feature importance graphs that help you explain model predictions and produces reports which can be used to support internal presentations or to identify issues with your model that you can take steps to correct.
Secure your data and code throughout the ML lifecycle
Amazon SageMaker offers a comprehensive set of security features, including encryption, private network connectivity, authorization, authentication, monitoring, and auditability to help your organization with security requirements that may apply to machine learning workloads.
Essential features for ML in production
Automate machine learning workflows
Amazon SageMaker Pipelines is the first purpose-built, easy-to-use continuous integration and continuous delivery (CI/CD) service for machine learning. Workflows can be shared and re-used between teams.
Improve transparency
Amazon SageMaker Clarify provides bias detection across the ML workflow, enabling you to build greater fairness and transparency into your ML model. SageMaker Clarify also includes feature importance graphs that help you explain model predictions and produces reports which can be used to support internal presentations or to identify issues with your model that you can take steps to correct.
Secure your data and code throughout the ML lifecycle
Amazon SageMaker offers a comprehensive set of security features, including encryption, private network connectivity, authorization, authentication, monitoring, and auditability to help your organization with security requirements that may apply to machine learning workloads.
Amazon SageMaker customers
Amazon SageMaker is used by tens of thousands of customers across a wide range of industries.
Get started with Amazon SageMaker JumpStart
Amazon SageMaker is a machine learning service that you can use to build, train, and deploy ML models for virtually any use case. For a quick technical introduction, see the SageMaker step-by-step guide. To help you get started with your ML project, Amazon SageMaker JumpStart offers a set of pre-built solutions for the most common use cases that you can deploy with just a few clicks. These solutions are fully customizable so you can modify them to suit the needs of your specific use case and datasets.
Predictive maintenance
Georgia Pacific uses SageMaker to develop ML models that detect machine issues early.
Computer vision
3M is using defect detection models built on SageMaker to improve the effectiveness of its quality control processes.
Autonomous driving
Lyft Level 5 standardized on SageMaker for training and reduced model training times from days to under a couple of hours.