SageMaker Unified Studio updates
SageMaker Unified Studio adds resource blueprints, catalog-scale data quality profiling with email notifications, consolidated metadata management, and DuckDB 2.0 plus extensions in notebooks.
The Deploy Log | Products | Free, no signup
24 updates to Amazon SageMaker from Amazon Web Services in The Deploy Log: each row Amazon Web Services published whose subject opens with Amazon SageMaker or SageMaker or names a feature in it, or names SageMaker anywhere in it. By publisher's dates alone: from to , a new one every day on the median, the longest wait was 29 days, and 19 updates in the last 30 days. The newest is SageMaker Unified Studio updates. 1 row with no publisher's date, each listed by its edition's date. Each links the publisher's own page.
SageMaker Unified Studio adds resource blueprints, catalog-scale data quality profiling with email notifications, consolidated metadata management, and DuckDB 2.0 plus extensions in notebooks.
Full Fine-Tuning is now supported on SageMaker AI Serverless Model Customization with no GPU provisioning or cluster setup.
SageMaker AI now monitors generative AI inference endpoints by default, notifying customers about slow responses, idle GPUs, and configuration gaps in six regions.
SageMaker AI now monitors generative AI inference endpoints and reports latency drift, unused GPU capacity, and configuration gaps by default.
Enhanced observability for SageMaker HyperPod diagnoses GPU training issues across EKS and Slurm.
Cross-Region inference routing is available for real-time inference in 32 AWS Regions and asynchronous inference in 25 Regions.
Higher pod density per node on EKS-orchestrated SageMaker HyperPod clusters, available in all HyperPod regions.
OpenTSLM time-series language models are now available in Amazon SageMaker JumpStart.
Amazon SageMaker AI adds auto-labeling, synthetic data generation, dataset assessment, and expert feedback via Python SDK APIs.
Amazon SageMaker AI real-time endpoints now support KV cache aware routing, enabled via an environment variable and RoutingStrategy set to KV_AWARE.
SageMaker HyperPod Inference now supports NVIDIA Dynamo for disaggregated and multi-node inference across 16 regions.
Cluster operations now track scaling and on-demand deep health check requests from submission to completion, at no additional cost.
KV cache-aware routing for real-time inference endpoints is available today in all AWS Regions where SageMaker AI is available.
SageMaker Inference supports Multi-Instance GPU to maximize GPU utilization for model endpoints.
Amazon SageMaker AI Studio now supports IAM Identity Center multi-region replication.
Amazon SageMaker AI real-time endpoints now support GPU partitioning.
Amazon SageMaker AI now supports SOCI-indexed images to start training jobs in seconds.
SageMaker MLflow Apps now support MLflow 3.16.
Amazon SageMaker AI now supports serverless model customization for Muse Glimmer, Qwen 3.8, and Ministral 3 models.
Amazon SageMaker AI now supports custom sequence lengths for serverless model customization jobs.
Reflection's Beam model is now available on Amazon SageMaker JumpStart and AWS Trainium.
Muse Spark Open model is now available on Amazon SageMaker JumpStart.
Amazon SageMaker HyperPod Inference Gateway is a Kubernetes-native, GPU-aware routing system that deploys as a single EKS managed addon and reduces first-token latency by up to 82%.
Amazon SageMaker AI now supports serverless model customization for NVIDIA Nemotron 3.5 Lightning.
One line every time AI actually ships, around the clock, free on Threads and Bluesky. Every one is in The Deploy Log with its source.