Amazon Web Services Inc. has released an improved version of its SageMaker Studio, a development interface that provides access to purpose-built artificial intelligence and machine learning development tools from preparing data, building training, deploying and managing machine learning models.

Announced today during re:Invent 2023, the company’s annual conference in Law Vegas, the new version includes a new web-based interface that loads faster and provides access to developers in their preferred integrated development environment, or IDE, to SageMaker’s machine learning tooling resources. An IDE is a software application used by developers to write, build and debug software.

SageMaker Studio was initially released in 2019 as a managed toolkit for accessing development resources and tools including access to JupyterLab, a popular AI development platform, and RStudio an IDE for the programming language R, which was designed for statistical computing. It allows developers to experiment with models, visualize data changes and perform debugging all in a single interface.

The new web-based interface for Studio serves as a central point for launching all of SageMaker’s tools allowing machine learning developers to build, train, fine-tune and deploy models. It’s now possible to view every stage of development from the web interface and access foundation models through SageMaker Jumpstart. Amazon added that with the web interface, there is no more need to manually upgrade Studio.

Additionally, Studio now includes a Code Editor that can be launched directly from the browser. With Code Editor, developers gain access to thousands of Visual Studio Code compatible extensions from the Open VSX registry and the preconfigured AWS toolkit for VS Code for deploying apps on AWS.

From within Code Editor, users will be able to use Amazon’s AI code suggester and security scanner powered by CodeWhisperer and CodeGuru to speed their development time. Using these AI-powered tools, users will be able to write, explain and refactor their code as they go, making it easier than ever to get going on an AI or machine learning development project.

Both Code Editor and JupyterLab can be launched into private instances that only the user has access to. These are flexible workspaces that are designed to provide faster and more efficient coding environments while maintaining privacy and security for developers.

Within these environments, Amazon preconfigured popular machine learning frameworks and Python packages. Even with the staged instances, users can still collaborate with an improved experience using built-in Git integration to share and version their code or bring shared file storage using Amazon EFS to access filesystems between different users and across teams.

The web-based interface also includes a new administration console that allows individual users to set up and launch their own SageMaker Studio with default presets using only one click. For enterprise administrators, Amazon included step-by-step instructions to follow regarding connecting identity providers, provisioning access policies and setting up applications.

Image: AWS