There are a lot of components to machine learning workflows, many of which come with their own set of tools that exist separately. Switching between multiple different tools and user interfaces eats away your productivity and slows down the pace of machine learning development. Amazon SageMaker Studio provides a single unified interface for all the tools you need to take your models from experimentation to production and boost your productivity.
You can easily login to Amazon SageMaker Studio using the single sign-on enabled by AWS SSO. You can then use the Amazon SageMaker Autopilot to automatically generate models from your data, or spin-up the new SageMaker Notebooks (currently in preview) in seconds to start building your ML models and algorithms. Collaborating on notebooks with your peers is easy in SageMaker Studio. With a single click, you can share a link to a snapshot of your notebook that is captured with all its dependencies and configurations to reproduce your analysis and results. As you start experimenting with various model parameters and inputs, you can use the SageMaker Studio’s visual interface for Amazon SageMaker Experiments to easily browse, track and compare your machine learning experiments, helping you keep track of incremental improvements and best models. SageMaker Studio also provides access to real-time alerts provided by Amazon SageMaker Debugger to help you troubleshoot your models as they train, optimize training times, and improve model quality. Once the model is deployed, SageMaker Studio also enables you to monitor, visualize, and analyze drifts detected via Amazon SageMaker Model Monitor, enabling you to continuously track and improve the quality of your predictions.
Starting today, Amazon SageMaker Studio is generally available in US East (Ohio) AWS region at no additional charge, with additional regions coming soon. Because SageMaker Studio Notebooks is in preview, visual elements of SageMaker Studio may be impacted. To learn more about the Amazon SageMaker Studio, read the blog here or refer to the documentation to quickly get started.