Search⌘ K
AI Features
Home
Courses
Grokking the AWS Certified Machine Learning Engineer - Associate

SageMaker Studio and ML Environments

Explore the AWS SageMaker Studio ecosystem and its environments tailored for different ML personas and stages. Understand how Studio, Studio Classic, and Studio Lab differ, learn about Data Wrangler for data preparation, and see how Canvas enables no-code ML. Discover integration with tools like Git, CI/CD pipelines, and AWS services that support automated production ML workflows.

We'll cover the following...

SageMaker Studio is a unified, web-based integrated development environment for the entire machine learning life cycle on AWS. This lesson walks through the SageMaker Studio ecosystem, covering Studio, Studio Classic, Studio Lab, Data Wrangler, Canvas, and supported IDEs. Each component maps to a specific stage of the ML pipeline and serves a distinct persona, whether that persona is a data scientist iterating on notebooks, an ML engineer building production pipelines, or a business analyst generating predictions without code.

SageMaker components in ML lifecycle and persona alignment
SageMaker components in ML lifecycle and persona alignment
Note: The exam frequently tests your ability to distinguish SageMaker Studio (an ML life cycle hub) from AWS Glue (ETL and data cataloging). Studio handles ML-focused workflows, and Glue handles broader data integration tasks.

Different ML personas benefit from different environments within the SageMaker Studio ecosystem, and the exam expects you to match tools to use cases precisely. The sections that follow break down each component, its integration points, and the trade-offs that drive architectural decisions.

Studio environment options and trade-offs

Three distinct environments exist under the SageMaker Studio umbrella, each targeting different use cases across the ML life cycle.

SageMaker Studio is the latest-generation, fully integrated ML IDE. It supports JupyterLab SpacesDedicated, configurable compute environments within Studio that allow users to select specific instance types and persistent storage, enabling both individual and shared collaborative workflows. and integrates with SageMaker Pipelines for orchestration, SageMaker Experiments for tracking across pipelines and SDK-based workflows, and the Model Registry for model versioning. When a data scientist launches a JupyterLab Space, Studio provisions the requested instance, mounts an Amazon EBS volume for persistence, and connects the session to IAM-scoped permissions that govern access to S3 buckets, ECR repositories, and other AWS resources.

SageMaker Studio Classic is the legacy, notebook-focused predecessor. It provides a single-user JupyterServer experience without the Spaces abstraction. SageMaker Studio Classic remains available for backward compatibility, ...