Amazon SageMaker Feature Store now supports batch feature writes and record listing

Amazon SageMaker Feature Store now supports new capabilities for high-throughput feature ingestion, record discovery, and offline store cataloging.

• BatchWriteRecord enables writing multiple records across feature groups in a single request with lower latency and fewer API calls
• ListRecords allows data scientists to retrieve and browse record identifiers in a feature group without knowing each ID in advance
• Custom table naming for Glue and Iceberg tables in the offline store for better organization
• BatchWriteRecord supports time-to-live settings at record, request, and feature group levels
• Individual record failures don't fail the entire batch request
• Available in all AWS Regions where SageMaker Feature Store is available

These capabilities enable data scientists to ingest features at scale and manage feature store records without building custom tooling.