AWS analytics at re:Invent 2025: Unifying Data, AI, and governance at scale

This article summarizes AWS analytics announcements at re:Invent 2025, focusing on unifying data warehouses, lakes, and AI with Apache Iceberg as the foundation.

• SageMaker Unified Studio enables one-click onboarding with SQL, Python, Spark, and natural language in single interface
• SageMaker Data Agent generates intelligent execution plans from natural language descriptions
• Apache Iceberg catalog federation allows querying remote tables without copying data
• Amazon Redshift now writes directly to Iceberg tables for open lakehouse architectures
• SageMaker Catalog exports metadata as queryable Iceberg tables for governance analysis
• Amazon Redshift federated permissions enforce consistent governance across multiple warehouses
• Amazon OpenSearch Service supports agentic search and GPU-accelerated vector database building
• Amazon EMR Serverless reduces costs 20% with serverless storage and auto-scaling
• AWS Glue materialized views accelerate queries up to 8x with automatic refresh
• Apache Spark upgrade agent transforms version upgrades from months to weeks
• Amazon MWAA Serverless eliminates operational overhead for workflow orchestration

AWS is fundamentally shifting analytics by unifying data platforms, simplifying access through AI, and automating governance to reduce operational overhead while improving performance and cost efficiency.