See how uniopen, a retail platform from Taiwan's Uni-President Enterprises Group, adapted Amazon Nova 2 Lite to its content-moderation policies using supervised fine-tuning in Amazon SageMaker AI and prompt optimization. Business-relevant evaluation and release gates kept quality in check.

Today, AWS announces a new bi-directional, zero-copy integration between AWS Analytics services and Workday Data Cloud. Instead of building and maintaining ETL pipelines to replicate data between systems, joint customers can now query across their AWS and Workday datasets in place, without moving or copying data, while retaining existing security controls in both environments. From AWS, you can use Amazon SageMaker Unified Studio, Amazon EMR, AWS Glue, Amazon Redshift, and Amazon Athena to query Workday data through Workday Live Data Query or federate to Apache Iceberg tables in Workday Data Cloud through catalog federation. Configure a connection in SageMaker Unified Studio, AWS Glue, or AWS Lake Formation consoles and begin querying immediately. Access is governed by AWS and Workday permissions. From Workday, you can register your AWS Glue Data Catalog Iceberg REST endpoint as an external catalog in Workday Analytics and query your AWS Iceberg tables directly. Common use cases include joining workforce data with business data for talent analytics, workforce planning, and machine learning. This integration is available in all AWS commercial regions where AWS Glue is available. To learn more, see the AWS Glue User Guide, the Workday Data Cloud integration with AWS on Workday Data Cloud, and the AWS Big Data Blog post.

Today, AWS announces a new bi-directional, zero-copy integration between AWS Analytics services and Workday Data Cloud. Instead of building and maintaining ETL pipelines to replicate data between systems, joint customers can now query across their AWS and Workday datasets in place, without moving or copying data, while retaining existing security controls in both environments. From AWS, you can use Amazon SageMaker Unified Studio, Amazon EMR, AWS Glue, Amazon Redshift, and Amazon Athena to query Workday data through Workday Live Data Query or federate to Apache Iceberg tables in Workday Data Cloud through catalog federation. Configure a connection in SageMaker Unified Studio, AWS Glue, or AWS Lake Formation consoles and begin querying immediately. Access is governed by AWS and Workday permissions. From Workday, you can register your AWS Glue Data Catalog Iceberg REST endpoint as an external catalog in Workday Analytics and query your AWS Iceberg tables directly. Common use cases include joining workforce data with business data for talent analytics, workforce planning, and machine learning. This integration is available in all AWS commercial regions where AWS Glue is available. To learn more, see the https://docs.aws.amazon.com/glue/latest/dg/connecting-to-data-workday.html, the https://doc.workday.com/admin-guide/en-us/workday-data-cloud/data-out/workday-data-lake/workday-data-lake.html, and the https://aws.amazon.com/blogs/big-data/introducing-the-workday-data-cloud-integration-with-aws/.

SSMとAWS GPU基盤で音声AIの低遅延化を解説

Amazon SageMaker Unified Studio now supports custom designations, named permission sets that control what actions individual project members can perform. Previously, project members could only be assigned Owner or Contributor designations, which provided nearly identical permissions. With custom designations, administrators define permission sets that align with their organization's personas and assign them to project members. Designations govern all actions within a project, providing administrators with a single permission model for separation of duties. Administrators create designations in the SageMaker Unified Studio administration portal by selecting permissions grouped into categories such as Catalog, Subscriptions, Connections, Notebooks, and Data Agent. For example, a "Data Steward" designation might include permissions to approve subscription requests and publish assets, while an "Analyst" designation is limited to querying data and running notebooks. Each designation is defined at the domain level and assigned across any project in that domain. When a member performs an action, SageMaker Unified Studio evaluates their assigned designation and permits or denies the request. Data engineers, stewards, analysts, and automated pipelines can now work within the same project with differentiated access, removing the need to create separate projects to isolate permissions. Every domain includes two managed designations, Owner and Contributor, and administrators can create up to 10 additional designations. This feature is available in all AWS Regions where Amazon SageMaker Unified Studio is available. To learn more, visit the Designations documentation.

Amazon SageMaker Unified Studio now supports custom designations, named permission sets that control what actions individual project members can perform. Previously, project members could only be assigned Owner or Contributor designations, which provided nearly identical permissions. With custom designations, administrators define permission sets that align with their organization's personas and assign them to project members. Designations govern all actions within a project, providing administrators with a single permission model for separation of duties. Administrators create designations in the SageMaker Unified Studio administration portal by selecting permissions grouped into categories such as Catalog, Subscriptions, Connections, Notebooks, and Data Agent. For example, a "Data Steward" designation might include permissions to approve subscription requests and publish assets, while an "Analyst" designation is limited to querying data and running notebooks. Each designation is defined at the domain level and assigned across any project in that domain. When a member performs an action, SageMaker Unified Studio evaluates their assigned designation and permits or denies the request. Data engineers, stewards, analysts, and automated pipelines can now work within the same project with differentiated access, removing the need to create separate projects to isolate permissions. Every domain includes two managed designations, Owner and Contributor, and administrators can create up to 10 additional designations. This feature is available in all https://docs.aws.amazon.com/sagemaker-unified-studio/latest/adminguide/supported-regions.html where Amazon SageMaker Unified Studio is available. To learn more, visit the https://docs.aws.amazon.com/sagemaker-unified-studio/latest/adminguide/designations.html documentation.

Amazon SageMaker AI inference now supports GPU partitioning, enabling you to run several models on a single GPU instead of dedicating a whole accelerator to each one. Using NVIDIA Multi-Instance GPU (MIG) technology, SageMaker AI divides a GPU into as many as seven isolated partitions, each with its own memory and compute. Customers serving small and mid-size models, such as fine-tuned variants, embedding models, rerankers, and classifiers, previously had to give every model an entire accelerator, so an instance hosted no more models than it had GPUs and paid for GPU capacity those models could never use. With GPU partitioning, you choose how to divide the GPUs behind your endpoint and place each model on a partition sized to fit it. Several models then share one accelerator with hardware-enforced isolation, so a busy model cannot take memory or throughput from the models beside it. Right-sizing an accelerator to a workload also opens up newer instance generations that would otherwise be more capacity than a single model needs. You configure GPU partitioning when you create an endpoint that hosts inference components, and you continue to invoke each model by name, so your application code doesn't change. GPU partitioning is available on P4d, P5, G7, and G7e instances in US East (N. Virginia), US East (Ohio), US West (Oregon), Asia Pacific (Singapore), Asia Pacific (Tokyo), Europe (Frankfurt), and Europe (Ireland). To learn more, visit the Amazon SageMaker AI deployment page and see the Amazon SageMaker AI Developer Guide.

Today, Amazon SageMaker AI announces support for Seekable OCI (SOCI)-indexed container images for training and processing jobs, so your jobs start running your code sooner. SOCI-indexed images start Amazon SageMaker AI training containers in seconds, cutting GPU idle time during image pulls by up to 90%, with no impact on training performance. Before a training or processing job can run, SageMaker AI downloads its container image to the instance. For deep learning images, which routinely exceed 10 GB, that download adds minutes to the start of every job, and you pay for that time. Now, when your image includes a SOCI index, SageMaker AI lazy-loads the image, fetching only the files needed to start the container and streaming the rest in the background as your code runs. To use this feature, reference a SOCI-indexed image from a private Amazon ECR repository in your training or processing job. You can use a pre-indexed AWS Deep Learning Container by changing the image tag to its -soci variant, or add a SOCI index to your own image in Amazon ECR with open-source tools, without rebuilding the image. No changes to your job definition, entry point, or data channels are needed. SageMaker AI checks for an index automatically and images without an index download as before. If an index can't be used, the job falls back to a standard download and continues running. SOCI-indexed images for SageMaker AI training and processing jobs are available today in all AWS Regions where SageMaker AI is available, through the SageMaker AI APIs, CLIs, SDKs, and Console. To learn more, see the documentation for https://docs.aws.amazon.com/sagemaker/latest/dg/train-soci-indexed-images.html and https://docs.aws.amazon.com/sagemaker/latest/dg/processing-job-soci-indexed-images.html, or read the https://aws.amazon.com/blogs/machine-learning/announcing-soci-indexed-images-for-faster-sagemaker-ai-jobs/.

Amazon SageMaker AI inference now supports GPU partitioning, enabling you to run several models on a single GPU instead of dedicating a whole accelerator to each one. Using NVIDIA Multi-Instance GPU (MIG) technology, SageMaker AI divides a GPU into as many as seven isolated partitions, each with its own memory and compute. Customers serving small and mid-size models, such as fine-tuned variants, embedding models, rerankers, and classifiers, previously had to give every model an entire accelerator, so an instance hosted no more models than it had GPUs and paid for GPU capacity those models could never use. With GPU partitioning, you choose how to divide the GPUs behind your endpoint and place each model on a partition sized to fit it. Several models then share one accelerator with hardware-enforced isolation, so a busy model cannot take memory or throughput from the models beside it. Right-sizing an accelerator to a workload also opens up newer instance generations that would otherwise be more capacity than a single model needs. You configure GPU partitioning when you create an endpoint that hosts inference components, and you continue to invoke each model by name, so your application code doesn't change. GPU partitioning is available on P4d, P5, G7, and G7e instances in US East (N. Virginia), US East (Ohio), US West (Oregon), Asia Pacific (Singapore), Asia Pacific (Tokyo), Europe (Frankfurt), and Europe (Ireland). To learn more, visit the https://aws.amazon.com/sagemaker/ai/deploy/ and see the https://docs.aws.amazon.com/sagemaker/latest/dg/inference-components-accelerator-partitioning.html.

Today, Amazon SageMaker AI announces support for Seekable OCI (SOCI)-indexed container images for training and processing jobs, so your jobs start running your code sooner. SOCI-indexed images start Amazon SageMaker AI training containers in seconds, cutting GPU idle time during image pulls by up to 90%, with no impact on training performance. Before a training or processing job can run, SageMaker AI downloads its container image to the instance. For deep learning images, which routinely exceed 10 GB, that download adds minutes to the start of every job, and you pay for that time. Now, when your image includes a SOCI index, SageMaker AI lazy-loads the image, fetching only the files needed to start the container and streaming the rest in the background as your code runs. To use this feature, reference a SOCI-indexed image from a private Amazon ECR repository in your training or processing job. You can use a pre-indexed AWS Deep Learning Container by changing the image tag to its -soci variant, or add a SOCI index to your own image in Amazon ECR with open-source tools, without rebuilding the image. No changes to your job definition, entry point, or data channels are needed. SageMaker AI checks for an index automatically and images without an index download as before. If an index can't be used, the job falls back to a standard download and continues running. SOCI-indexed images for SageMaker AI training and processing jobs are available today in all AWS Regions where SageMaker AI is available, through the SageMaker AI APIs, CLIs, SDKs, and Console. To learn more, see the documentation for training jobs and processing jobs, or read the launch blog.

Amazon SageMaker AI MLflow Apps now support MLflow 3.16, bringing a redesigned tracing experience and faster experiment tracking to your generative AI and machine learning workflows. MLflow 3.16 improves debugging and analysis of generative AI applications and agents. The redesigned trace explorer is now the default, with adjustable row height, reorderable columns, custom columns from any trace tag or metadata field. This release also accelerates metric charts for experiments with large metric histories. New MLflow Apps are created on MLflow 3.16 by default. Existing MLflow Apps will be upgraded to MLflow 3.16 during an upcoming maintenance window, with no action required. MLflow 3.16 is available in all AWS Regions where MLflow Apps are available. To learn more, visit the SageMaker AI MLflow developer guide.

Amazon SageMaker AI MLflow Apps now support MLflow 3.16, bringing a redesigned tracing experience and faster experiment tracking to your generative AI and machine learning workflows. MLflow 3.16 improves debugging and analysis of generative AI applications and agents. The redesigned trace explorer is now the default, with adjustable row height, reorderable columns, custom columns from any trace tag or metadata field. This release also accelerates metric charts for experiments with large metric histories. New MLflow Apps are created on MLflow 3.16 by default. Existing MLflow Apps will be upgraded to MLflow 3.16 during an upcoming maintenance window, with no action required. MLflow 3.16 is available in all AWS Regions where MLflow Apps are available. To learn more, visit the https://docs.aws.amazon.com/sagemaker/latest/dg/mlflow.html.

Today, AWS announces light mode support in Amazon SageMaker Unified Studio for Identity Center (IDC) based domains. Customers can now configure the visual interface mode to match their preference, choosing between dark and light themes. With this launch, customers can set this visual preference in both IAM-based and IDC-based domains.  Light mode helps improve readability in bright environments and provides a familiar visual experience for customers who prefer lighter interfaces. Combined with the existing dark mode, this update gives you full control over the appearance of your development environment, improving accessibility and reducing eye strain across varying lighting conditions. In SageMaker Unified Studio settings, you can choose 'Customize appearance' under your Profile settings to select between visual modes including dark mode, light mode, and device setting, which matches your operating system's setting. The setting persists across browsers and devices. This feature is available in all AWS Regions where Amazon SageMaker Unified Studio is available. To learn more, see the https://docs.aws.amazon.com/sagemaker-unified-studio/latest/userguide/navigating-sagemaker-unified-studio.html#display-mode

Today, AWS announces light mode support in Amazon SageMaker Unified Studio for Identity Center (IDC) based domains. Customers can now configure the visual interface mode to match their preference, choosing between dark and light themes. With this launch, customers can set this visual preference in both IAM-based and IDC-based domains.  Light mode helps improve readability in bright environments and provides a familiar visual experience for customers who prefer lighter interfaces. Combined with the existing dark mode, this update gives you full control over the appearance of your development environment, improving accessibility and reducing eye strain across varying lighting conditions. In SageMaker Unified Studio settings, you can choose 'Customize appearance' under your Profile settings to select between visual modes including dark mode, light mode, and device setting, which matches your operating system's setting. The setting persists across browsers and devices. This feature is available in all AWS Regions where Amazon SageMaker Unified Studio is available. To learn more, see the Amazon SageMaker Unified Studio User Guide.

GPUインフラを管理せずカスタム学習ループを実装可能なSageMakerの新APIを解説。

Amazon SageMaker AI serverless model customization now supports training session APIs, giving customers full control over their post-training algorithm without provisioning or managing GPU infrastructure. Managed techniques, including SFT, DPO, and other reinforcement learning methods, let customers pick a technique and a model and have SageMaker AI run the training. With training session APIs, teams get to write the training loop, whether that's reproducing SFT or DPO on their own terms or implementing a custom loss, a new RL reward signal, a data curriculum, or an algorithm straight from the latest research. Operations such as forward_backward(), optim_step(), and sample() become API calls against a live, stateful session, so customers compose whatever training procedure their technique requires while SageMaker AI handles GPU allocation, model and LoRA loading, distributed execution, fault recovery, checkpoint storage, and security. Customers can call these operations directly from their own Python code to drive reinforcement learning, supervised fine-tuning, or custom training objectives against a live session, with checkpointing and resume built in. Billing is per token processed, so local computation like reward scoring isn't billed as GPU time. SageMaker AI is also releasing the SageMaker training sessions cookbook, a set of ready-to-run SFT and RL recipes customers can use as blueprints for their own training loops. Training session APIs are generally available today in US East (N. Virginia), US West (Oregon), and Asia Pacific (Mumbai). To get started, install the SageMaker Python SDK and clone the cookbook, or launch a notebook in SageMaker Studio and connect a coding agent. To learn more, visit the documentation and the Amazon SageMaker AI webpage.

Amazon SageMaker AI serverless model customization now supports training session APIs, giving customers full control over their post-training algorithm without provisioning or managing GPU infrastructure. Managed techniques, including SFT, DPO, and other reinforcement learning methods, let customers pick a technique and a model and have SageMaker AI run the training. With training session APIs, teams get to write the training loop, whether that's reproducing SFT or DPO on their own terms or implementing a custom loss, a new RL reward signal, a data curriculum, or an algorithm straight from the latest research. Operations such as forward_backward(), optim_step(), and sample() become API calls against a live, stateful session, so customers compose whatever training procedure their technique requires while SageMaker AI handles GPU allocation, model and LoRA loading, distributed execution, fault recovery, checkpoint storage, and security. Customers can call these operations directly from their own Python code to drive reinforcement learning, supervised fine-tuning, or custom training objectives against a live session, with checkpointing and resume built in. Billing is per token processed, so local computation like reward scoring isn't billed as GPU time. SageMaker AI is also releasing the https://github.com/aws/sagemaker-training-sessions-cookbook, a set of ready-to-run SFT and RL recipes customers can use as blueprints for their own training loops.https://docs.aws.amazon.com/sagemaker/latest/dg/customizing-models-training-sessions.html Training session APIs are generally available today in US East (N. Virginia), US West (Oregon), and Asia Pacific (Mumbai). To get started, install the SageMaker Python SDK and clone the cookbook, or launch a notebook in SageMaker Studio and connect a coding agent. To learn more, visit the  https://docs.aws.amazon.com/sagemaker/latest/dg/customizing-models-training-sessions.htmland the https://aws.amazon.com/sagemaker/ai/model-customization/

Amazon SageMaker AI now supports serverless model customization for Muse Glimmer, Qwen3.8 27B, and Ministral 3 3B, 8B, and 14B models. In addition to deploying these models on SageMaker AI, you can now adapt them to your specific domains and workflows, using supervised fine-tuning (SFT), direct preference optimization (DPO), and reinforcement fine-tuning (RFT) techniques. You can also adapt these models with either low-rank adaptation (LoRA) or full fine-tuning techniques. This launch adds these models to a wide variety of models available for serverless customization on SageMaker AI, including models from the Nova, Gemma 4, Nemotron 3.5, Qwen, Llama, gpt-oss, and DeepSeek families. Model customization lets you tailor foundation models with your proprietary data so a smaller, right-sized model can match frontier-model quality on your tasks, while also reducing cost and latency. You can use labeled data with SFT to improve accuracy on domain-specific tasks, preference data with DPO to align outputs with your organization's tone, or reward signals with RFT to improve performance on new tasks. With serverless customization, SageMaker AI handles all infrastructure provisioning and training orchestration, so you can focus on your data and evaluation rather than cluster management, and only pay for what you use. Serverless model customization on SageMaker AI is available in US East (N. Virginia), US West (Oregon), Asia Pacific (Tokyo), and Europe (Ireland). To get started, navigate to the Models page in Amazon SageMaker Studio to launch a customization job, use the https://sagemaker.readthedocs.io/en/stable/model_customization/index.html or use our https://docs.aws.amazon.com/sagemaker/latest/dg/model-customize-agent-skills.html with your choice of coding assistant. To learn more and see the full supported list of models, see the Amazon SageMaker AI model customization https://docs.aws.amazon.com/sagemaker/latest/dg/customizing-models.html#customizing-models-supported.

Amazon SageMaker AI now supports serverless model customization for Muse Glimmer, Qwen3.8 27B, and Ministral 3 3B, 8B, and 14B models. In addition to deploying these models on SageMaker AI, you can now adapt them to your specific domains and workflows, using supervised fine-tuning (SFT), direct preference optimization (DPO), and reinforcement fine-tuning (RFT) techniques. You can also adapt these models with either low-rank adaptation (LoRA) or full fine-tuning techniques. This launch adds these models to a wide variety of models available for serverless customization on SageMaker AI, including models from the Nova, Gemma 4, Nemotron 3.5, Qwen, Llama, gpt-oss, and DeepSeek families. Model customization lets you tailor foundation models with your proprietary data so a smaller, right-sized model can match frontier-model quality on your tasks, while also reducing cost and latency. You can use labeled data with SFT to improve accuracy on domain-specific tasks, preference data with DPO to align outputs with your organization's tone, or reward signals with RFT to improve performance on new tasks. With serverless customization, SageMaker AI handles all infrastructure provisioning and training orchestration, so you can focus on your data and evaluation rather than cluster management, and only pay for what you use. Serverless model customization on SageMaker AI is available in US East (N. Virginia), US West (Oregon), Asia Pacific (Tokyo), and Europe (Ireland). To get started, navigate to the Models page in Amazon SageMaker Studio to launch a customization job, use the SageMaker Python SDK or use our Agent skills with your choice of coding assistant. To learn more and see the full supported list of models, see the Amazon SageMaker AI model customization documentation.

Today, Amazon SageMaker AI announces the ability to configure the sequence length used during your serverless model customization job. Amazon SageMaker AI serverless model customization supports fine-tuning for many open-weight models, including models from popular families including gpt-oss, Gemma, Llama, Nemotron, and Qwen. Previously, each fine-tuning recipe ran at a fixed sequence length that customers could not modify. Customers sometimes need to modify these parameters to achieve better results for their use case. Customers also asked for longer sequence lengths than the static defaults we supported on some models. Now, you can set sequence length parameters as part of the customization job's hyperparameters, anywhere up to the published maximum for your chosen model and technique. You can set the dataset_max_len hyperparameter for SFT and DPO jobs or set max_prompt_length and max_response_length hyperparameters for RFT jobs. This gives you more control to match your training data characteristics, whether you are working with short question-answer pairs or multi-page documents. Published maximums are up to 16K tokens for many models depending on model and technique. Serverless model customization on SageMaker AI is available in US East (N. Virginia), US West (Oregon), Asia Pacific (Tokyo), and Europe (Ireland). To get started, navigate to the Models page in Amazon SageMaker Studio to launch a customization job, use the SageMaker Python SDK or use our Agent skills with your choice of coding assistant. To learn more, and see the list of models and supported sequence lengths per technique, see the SageMaker AI documentation.