Amazon S3 Tables now support all Apache Iceberg V3 data types, including variant, nanosecond timestamps, geometry, geography, deletion vectors, and row lineage for improved analytics performance.

AWS Parallel Computing Service now supports scaling logs to help troubleshoot compute node scaling issues across CloudWatch Logs, S3, and Firehose.

AWS announces general availability of AMD-based R8a and M8a database instances powered by 5th generation AMD EPYC processors for Amazon Aurora and RDS.

Amazon S3 Tables now supports all Apache Iceberg V3 data types including geometry, geography, nanosecond timestamps, and column default values.

Amazon Managed Grafana now supports Grafana 13.2 workspaces with Git Sync, dynamic dashboards, and PromQL query support in CloudWatch.

This article demonstrates how to automate cross-partition AWS GovCloud (US) account bootstrapping using API Gateway, Lambda, and Parameter Store to securely provision accounts across partition boundaries.

An open-source AWS solution using Amazon Bedrock and Step Functions automates biomedical literature analysis, reducing manual paper review from 3-5 days to under 5 minutes.

AWS IAM Identity Center Identity Store APIs now accept Amazon Resource Names (ARNs) in addition to resource IDs, simplifying application code and eliminating parsing errors.

AWS CLI now supports bulk skill updates and version checks for the Agent Toolkit, allowing developers to update all outdated skills at once instead of individually.

AWS releases the Dogwood Local Engine, an open source library for enforcing temporal governance policies that regulate AI agent tool calls based on past actions and their outcomes.

AWS accounts now support phone number verification via SMS OTP for primary contact phone numbers, with inheritance support for AWS Organizations.

Amazon S3 Vectors now supports metadata pre-filtering to improve recall by up to 5x on filtered searches, plus a new prefix match operator for path and URL filtering.

Amazon S3 Vectors now supports metadata pre-filtering to improve recall on filtered searches by evaluating filters before similarity search, with no additional cost or re-ingestion required.

Amazon Aurora Serverless now scales faster, adding up to 16 ACUs per second and supporting up to 256 ACUs, making it ideal for agentic AI and bursty workloads.

AWS AppConfig experimentation enables A/B testing in production by exposing feature changes to controlled traffic slices and measuring real-world impact before full rollout.

AWS Transfer Family now automatically approves SFTP connector quota increases up to 1,000 connectors per account per region without manual approval.

Amazon RDS for MySQL now supports MySQL 26.7 in the Database Preview Environment for pre-release evaluation.

AWS announces UltraFast mode for OpenAI's GPT-6 Astra on Amazon Bedrock, delivering up to 6x faster inference with 300 tokens per second for latency-sensitive applications.

Amazon RDS now supports M8a database instances powered by 5th generation AMD EPYC processors for PostgreSQL, MySQL, and MariaDB workloads.

Amazon Aurora and RDS now support R8a database instances powered by 5th generation AMD EPYC processors with up to 75 Gbps network bandwidth.

Partner Revenue Measurement now supports multi-Partner resource tagging, allowing multiple AWS Partners to receive revenue attribution for the same resource.

This article demonstrates how to conduct multi-day Availability Zone evacuation drills using AWS ARC Zonal Shift to validate sustained resilience across ECS, EKS, RDS, and Aurora databases.

AWS Deadline Cloud now supports expressions and rich parameter types in job templates, enabling customers to express pipeline logic directly using Python-style syntax.

AWS Deadline Cloud now supports running Docker containers from Amazon ECS on Linux-based service-managed fleets for compute-intensive workloads.

MHK built SmartProminence AI Orchestrator, a HIPAA-eligible agentic workflow framework on Amazon Bedrock that reduced manual medical review effort by 90% and accelerated AI feature deployment from 3-6 months to 2 weeks.

Amazon CloudWatch Logs Insights now provides byte-scan estimates before query execution, helping users optimize log group selection and time ranges without incurring charges.

Kiro Crew enables self-learning agents for telecom network operations by preserving human corrections as reusable diagnostic lessons and converting demonstrated procedures into governed skills.

Amazon CloudWatch Logs now automatically indexes frequently queried fields to speed up Insights queries without manual configuration.

VPC Flow Logs Version 11 now supports embedding EC2 resource tags and next-hop interface metadata directly in flow log records, eliminating the need for custom enrichment pipelines.

Aurora PostgreSQL now supports direct querying of Apache Iceberg and Parquet data in S3 and AWS Glue without ETL pipelines or data duplication.

AWS Marketplace launches an AI agent skill that helps sellers build and validate usage-based metering integrations for SaaS products with guided setup and automated testing.

Amazon Aurora PostgreSQL now supports direct querying of Apache Iceberg and Parquet data in data lakes without ETL pipelines, using embedded DuckDB technology.

AWS launches three new Jam experiences for building, deploying, and hardening AI agents using Amazon Bedrock and related AWS services.

AWS announces three new AWS Heroes for September 2026: Avinash Shashikant Dalvi (serverless expert from India), Joanne Skiles (engineering leader from USA), and Xiaofei Li (community leader from China).

Resolve PostgreSQL replication lag in CDC migrations by implementing heartbeat tables to prevent WAL accumulation and keep replication slots advancing.

This article demonstrates how to migrate IBM Db2 z/OS databases to Amazon Aurora PostgreSQL using AWS DMS with an EC2-based Db2 gateway server for secure data replication and periodic refresh.

AWS CloudTrail introduces Event Coverage, a console feature that provides visibility into data plane operations logging across accounts and organizations, helping identify gaps in security monitoring.

Kiro, an agentic AI-powered development environment, accelerates AS/400 business rule extraction from months to days by autonomously analyzing RPG/COBOL code and generating modernization-ready documentation.

This article demonstrates how to build a claims assistant using Amazon Bedrock Knowledge Bases with agentic retrieval to answer natural-language questions about insurance claims with citations from source documents.

This article demonstrates how to build an AI-powered migration assessment agent using Amazon Bedrock AgentCore that automates evaluation of applications for Amazon EKS migration readiness.

Build a multi-agent music production pipeline on Amazon Bedrock AgentCore Runtime Instances that enables persistent, GPU-accelerated workflows with agents collaborating through shared sessions and volumes.

athenahealth modernized healthcare workloads using Amazon EKS Hybrid Nodes to achieve 50% faster response times, 50% cost reduction, and unified Kubernetes operations across on-premises and cloud environments while maintaining HIPAA compliance.

This article presents four architecture options for delivering Amazon Connect Customer conversational analytics transcripts to Amazon Redshift, ranging from fully managed (~15 min latency) to real-time streaming during calls (2-5 sec latency).

This article presents four architecture options for delivering Amazon Connect Customer conversational analytics transcripts to Amazon Redshift, ranging from ~15-minute managed latency to ~3-7 minute event-driven Lambda pipelines.

Amazon Connect Customer now offers native AI agents for email support that automatically summarize threads, provide policy-grounded guidance with citations, and draft replies for agent review before sending.

A voice-activated AI agent built with Amazon Connect, Amazon Bedrock, and Model Context Protocol enables supervisors to manage contact center AI agents through secure natural language commands instead of console access.

AWS launches Kiro, an AI-powered engineering agent that accelerates SAP modernization by automating code conversion and reducing migration effort by up to 87%.

AWS Kiro-based modernization agents accelerate SAP transformations through AI-powered automation of ABAP upgrades, code refactoring, integration migrations, and data warehouse modernization with significant effort reductions.

Amazon Bedrock now offers Anthropic's Claude models with in-country inferencing in India through geographic cross-Region inference between Mumbai and Hyderabad regions.

Amazon Bedrock now supports Anthropic Claude Opus 5 and Claude Sonnet 5 models with in-region inference in Seoul and Singapore for strict data residency requirements.