Run Positron on Amazon SageMaker AI for data science workflows
aws ·
Positron, an integrated development environment for data science, is now available on Amazon SageMaker AI, allowing teams to unify their workflows within a single browser-based interface. This integration enables data scientists to access governed AWS data sources like Amazon Athena and S3 directly through the Space execution role without managing separate credentials. The solution supports cross-language analysis by combining R for feature validation with Python for model training using XGBoost. Posit Assistant can leverage Amazon Bedrock for AI coding assistance while keeping data secure within the customer's account. The workflow demonstrated includes deploying a real-time endpoint via SageMaker AI and building a live inference application using Shiny for Python. Additionally, Quarto is used to generate reproducible reports that document the entire process from data discovery to deployment. Administrators configure this setup by registering a custom Positron container image in Amazon ECR and attaching it to a SageMaker Studio domain.
aws