Amazon SageMaker Feature Store now supports individual feature updates
Amazon SageMaker Feature Store introduces support for feature-level writes, allowing data scientists to update specific feature values within a record without rewriting the entire entry. This new capability reduces write latency and cost by enabling partial updates, streamlining data ingestion pipelines. It eliminates the need for manual merge logic when multiple pipelines write to the same feature group independently. This functionality is now generally available in all AWS Regions where SageMaker Feature Store is offered.
Features (1) ›
- Enable individual feature updates for records
SageMaker Feature Store now supports feature-level writes, allowing data scientists to update one or more feature values in a record with a single request. This capability means only the specified features are updated, leaving others unchanged, which significantly lowers write latency and cost. It simplifies data ingestion by removing the need for read-modify-write patterns or complex merge logic in pipelines.
https://aws.amazon.com/about-aws/whats-new/2026/08/sgm-feature-store-update-record/
Related releases
- Amazon SageMaker Batch Transform adds support for G6e instances AWS What's New ·
- Amazon SageMaker Unified Studio Workflows support Python and Bash operators AWS What's New ·
- SageMaker Unified Studio CI/CD adds notebook promotion and AI-assisted manifest generation AWS What's New ·
- Build Medallion Architectures with Iceberg MVs in Amazon SageMaker AWS Big Data Blog ·
- AWS Config expands support for 60 new resource types AWS What's New ·
- Meta's Muse-Glimmer-30B and Alibaba's Qwen 3.8-27B Models Now on SageMaker JumpStart AWS What's New ·