All news with #amazon sagemaker tag

AWS Glue Adds Zero-ETL Support for More SAP Entities

🔄 AWS Glue now provides full snapshot and incremental zero-ETL ingestion for additional SAP entities. The update adds snapshot ingestion for entities without deletion tracking and timestamp-based incremental loads for non-ODP systems, extending existing ODP support. Organizations can ingest SAP data directly into Amazon Redshift or the lakehouse architecture used by Amazon SageMaker, reducing engineering effort and operational complexity. This feature is available in all Regions where AWS Glue zero-ETL is offered.

Amazon SageMaker Studio Integrates EMR on EKS with SSO

🔒 Amazon SageMaker Unified Studio now supports EMR on EKS as a compute option for interactive Apache Spark sessions, bringing containerized, large-scale distributed compute with automatic scaling and cost optimizations directly into the Studio environment. The feature adds trusted identity propagation through AWS Identity Center, enabling single sign-on and end-to-end data access traceability for interactive analytics. Data practitioners can use corporate credentials to access Glue Data Catalog resources from SageMaker JupyterLab while administrators retain fine-grained access controls and audit trails. This capability is available in all existing SageMaker Unified Studio regions.

SageMaker Studio: Long‑Running Sessions with Corporate IDs

⏳ Amazon SageMaker Unified Studio now supports long-running background sessions using corporate identities via AWS IAM Identity Center's trusted identity propagation (TIP). Users can launch interactive notebooks and data processing on SageMaker, Amazon EMR, and AWS Glue that persist when they log off or experience network or credential interruptions. Sessions retain corporate permissions and can run up to 90 days (default 7 days), reducing the need for continuous monitoring and improving productivity for multi-hour or multi-day workloads.

Amazon SageMaker Catalog Enforces Glossary Metadata

📌 Amazon SageMaker Catalog now enforces glossary-term metadata during asset publishing. Administrators can require data producers to tag assets with approved business vocabulary from organizational glossaries, and enforcement rules will block publication if required terms are missing. This standardizes metadata, aligns technical schemas with business language, and improves discoverability and governance. Available in all regions where Amazon SageMaker Catalog operates; policies can be managed via the console, CLI, or SDKs.

Amazon SageMaker Catalog Adds Column-Level Metadata

📣 Amazon SageMaker Catalog now supports custom column-level metadata forms and markdown-enabled rich text descriptions so data stewards can attach business-specific key-value metadata and formatted documentation directly to individual columns. Form values and rich text are indexed in real time and become immediately searchable alongside column names, descriptions, and glossary terms. This capability is available in all AWS Regions where SageMaker is supported.

Amazon SageMaker Catalog Adds S3 Read/Write Access

📂 Amazon SageMaker Catalog now supports read and write access to Amazon S3 general purpose buckets, enabling data scientists and analysts to discover, process, and share unstructured data alongside structured datasets. Data publishers can grant read-only or read/write permissions when approving subscriptions or sharing S3 data, allowing processed outputs to be written back to the original bucket or folder. This feature is available in all Regions that support SageMaker Unified Studio and can be accessed via the studio UI, the Amazon DataZone API, SDK, or AWS CLI.

Amazon SageMaker Adds Custom Tags for Project Resources

🔖 Amazon SageMaker Unified Studio now lets administrators define custom tags that are applied to resources created by a SageMaker project. Administrators configure project profiles to supply tag key/value pairs or keys with default values that users can modify during project creation, helping enforce tagging standards and support SCPs and cost allocation. This initial release is API-only and available across all supported AWS Regions.

AWS Config Adds 52 New Resource Types Across Key Services

🔔 AWS Config now supports 52 additional AWS resource types across services including Amazon EC2, Amazon Bedrock, and Amazon SageMaker. With recording for all resource types enabled, AWS Config will automatically begin tracking these additions and they are available to Config rules and aggregators. You can monitor the new types in all Regions where supported, expanding discovery, assessment, audit, and remediation coverage.

SageMaker Unified Studio adds searchable match context

🔍 Amazon SageMaker in Unified Studio now surfaces additional search context that clarifies why each result appears by showing which metadata fields matched a query. Inline highlighting emphasizes matched terms and an explanation panel details matches across name, description, glossary, schema, and other metadata. The enhancement reduces time spent evaluating irrelevant assets by presenting match evidence directly in search results, enabling quicker validation without opening individual assets. The capability is available in all AWS Regions where SageMaker is supported.

SageMaker Studio Integrates with Athena Workgroups

📊 Data engineers and analysts can now connect Amazon SageMaker Unified Studio to existing Amazon Athena workgroups to run SQL queries using the workgroups' default settings and properties. This lets teams reuse access controls, cost limits, and query-tracking policies already defined in Athena, reducing setup time while maintaining governance. To enable it, choose 'Add compute' → 'Connect to existing compute resources' in Unified Studio; the connected Athena workgroup then appears in the query editor and is available in all regions where Unified Studio is supported.

Amazon SageMaker Notebooks Now Support Amazon Linux 2023

🆕 Amazon SageMaker notebook instances now support Amazon Linux 2023, giving data scientists and developers access to an updated, rpm-based runtime for managed Jupyter notebooks. AL2023 is the successor to AL2, offering a predictable two-year major release cadence and five years of long-term support. Enhanced security features include SELinux and FIPS 140-3 validated cryptographic modules. New notebook instances can be launched with either AL2023 or AL2.

Amazon SageMaker Notebooks Now Support Amazon Linux 2023

🚀 Amazon SageMaker notebook instances now offer Amazon Linux 2023 as a launch option alongside Amazon Linux 2. The update provides a modern rpm-based runtime with a predictable two-year release cycle and five years of long-term support. Enhanced security features include SELinux support and FIPS 140-3 validated cryptographic modules. Use AL2023 to benefit from updated packages and continued OS maintenance.

Amazon OpenSearch Service Adds Batch AI Inference Support

🧠 You can now run asynchronous batch AI inference inside Amazon OpenSearch Ingestion pipelines to enrich and ingest very large datasets for Amazon OpenSearch Service domains. The same AI connectors previously used for real-time calls to Amazon Bedrock, Amazon SageMaker, and third parties now support high-throughput, offline jobs. Batch inference is intended for offline enrichment scenarios—generating up to billions of vector embeddings—with improved performance and cost efficiency versus streaming inference. The feature is available in regions that support OpenSearch Ingestion on domains running 2.17+.

Amazon SageMaker Managed MLflow Now in AWS GovCloud

🛡️ Amazon SageMaker managed MLflow is now available in both AWS GovCloud (US-West) and AWS GovCloud (US-East) regions. The managed service integrates MLflow experiment tracking with SageMaker capabilities, streamlining AI experimentation and accelerating GenAI development from idea to production. It provides end-to-end observability to help reduce time-to-market and simplify compliance and operational oversight for government workloads.

Amazon DataZone Now Available in Three Additional Regions

🔔 Amazon DataZone is now available in AWS Asia Pacific (Hong Kong), Asia Pacific (Malaysia), and Europe (Zurich) Regions. The fully managed Amazon DataZone service catalogs, discovers, analyzes, shares, and governs organizational data, integrating with AWS Glue Data Catalog and Amazon Redshift. Consumers can search, subscribe, and analyze assets using tools like Amazon Redshift and Amazon Athena from the DataZone portal. The service also underpins governance in the next generation of Amazon SageMaker to simplify discovery and secure access to data and models.

Managed Tiered Checkpointing for Amazon SageMaker HyperPod

⚡ Amazon Web Services has announced general availability of managed tiered checkpointing for Amazon SageMaker HyperPod, a hybrid checkpointing capability that caches frequent checkpoints in CPU memory and periodically persists them to Amazon S3 for durability. The approach reduces model recovery time and minimizes training progress loss on large-scale clusters. It integrates with PyTorch Distributed Checkpoint (DCP) and is enabled via a CreateCluster/UpdateCluster API parameter; customers can use the sagemaker-checkpointing Python library to adopt it with minimal code changes. Currently available for HyperPod clusters using the EKS orchestrator.

Improved AI Assistance in Amazon SageMaker Unified Studio

🤖 Amazon Web Services announced enhancements to the Amazon Q Developer chat experience within SageMaker Unified Studio Jupyter notebooks and added a command-line interface for use in notebooks and the Code Editor. By integrating with Model Context Protocol (MCP) servers, the assistant becomes aware of project resources—data, compute, and code—and provides personalized, context-aware help. These updates aim to speed tasks like code refactoring, file edits, and troubleshooting while preserving transparency around assistant actions. The capabilities are available at no additional cost via the Amazon Q Developer Free Tier where SageMaker Unified Studio is offered; customers can enable Amazon Q Developer Pro for expanded functionality.

Amazon SageMaker Adds Account-Agnostic Project Profiles

🔁 Amazon SageMaker introduces account-agnostic, reusable project profiles within the SageMaker Unified Studio domain, enabling domain administrators to define project templates once and reuse them across multiple AWS accounts and regions. Profiles are decoupled from specific accounts and regions and can reference a new account pool for dynamic account and region selection at project creation, driven by custom authorization policies or predefined strategies. This reduces duplication, simplifies governance, and accelerates onboarding across large-scale data and ML environments. The feature is available in all Regions where Unified Studio is supported.

Amazon SageMaker Lakehouse Adds Tag-Based Access Control

🏷️ Amazon SageMaker lakehouse now supports tag-based access control (TBAC) across federated catalogs, extending capability beyond the default AWS Glue Data Catalog to Amazon S3 Tables, Amazon Redshift, and federated sources such as DynamoDB, PostgreSQL, and SQL Server. TBAC lets administrators group resources with tags, grant access based on those tags, and rely on tag inheritance so new tables automatically receive fine-grained controls. Administrators can create and apply tags via the AWS Lake Formation console and grant tag-based permissions to principals; tagged resources are then usable through Amazon Athena, Amazon Redshift, Amazon EMR, and SageMaker Unified Studio. The feature is available in all commercial AWS Regions via the Console, AWS CLI, and SDKs, with supporting Lake Formation Tags documentation and a blog post.

SageMaker HyperPod Supports Customer-Managed KMS for EBS

🔐 Amazon SageMaker HyperPod now supports customer-managed AWS KMS keys (CMKs) to encrypt EBS volumes, giving enterprises direct control over encryption for root and secondary storage. This enables integration with existing key management and compliance workflows and uses a grants-based approach for secure cross-account access. Customers can specify CMKs via the CreateCluster and UpdateCluster APIs for clusters in continuous provisioning mode. The capability is available in all Regions where HyperPod runs.