AWS News - 2026-10-01

Amazon Quick now supports the hierarchy filter, a single control that lets readers drill through related dimensions such as Region, Country, and City. It replaces several stacked filter controls with one, cutting clutter and guiding readers to their data in fewer clicks.

The filter holds up to five dimension levels, arranged broadest to most detailed. Readers open one dropdown and expand each level in turn, with every selection narrowing the next and auto-selecting its parent chain. Authors create it by adding a filter, setting its type to Hierarchy filter, and arranging the fields parent to child.

For a retail sales dashboard, instead of stacking Region, Country, and City controls side by side, an author adds one hierarchy filter. A reader expands a region to see its countries, then a country to see its cities, selecting any mix along the way. The relationships are visible in the control, so readers do not need to know which country belongs to which region.

The hierarchy filter is available today in all AWS Regions where Amazon Quick is supported. To see how it works, read the blog post. To learn more, see Hierarchy filters in the Amazon Quick User Guide.

Today, Amazon Quick adds support for building applications that use live data directly from your Quick datasets, so the KPIs, charts, tables, and insights in your application always reflect the latest data. These applications go beyond viewing data. They facilitate intelligent, data-driven business operations workflows where each person can act on what they see, from handling requests and approvals to triggering the next step. Each application user sees data under their own Amazon Quick permissions, so the row-level and column-level security you already set up is applied.

To build one, give the app builder chat agent the name of the dataset you want to use and describe in plain language what you want to build. This lets you enrich an application, such as a team portal that handles requests and approvals and now also shows current pipeline numbers, where each application user sees only the data their permissions allow.

This feature is available in all supported regions of apps in Amazon Quick. To see how it works, read the blog post. To learn more, see Connecting to datasets in the Amazon Quick User Guide.

 

AWS CloudTrail introduces Event Coverage, a new console experience that gives customers visibility into their data plane operations coverage at the account and organization level. You can now see which AWS services and resource types in your environment have data event logging enabled and which do not. This helps you quickly identify gaps in your logging posture without manually scanning accounts in your organization.

Data events enable you to track data plane operations in AWS services. For example, you can log Amazon S3 object-level operations like GetObject and PutObject to detect unauthorized data access or exfiltration attempts. Without comprehensive data events coverage, these activities can go unnoticed, leaving blind spots in your security monitoring. With Event Coverage, you can view coverage across all supported data event sources in one place and subscribe to data events directly from the dashboard. This makes it easier to close coverage gaps in a few clicks, especially for organizations managing multiple accounts where tracking coverage across services can be time-consuming.

You can access Event Coverage from the AWS CloudTrail console. This feature is available in all commercial AWS Regions where AWS CloudTrail is supported. To learn more about logging Data Events, visit the AWS CloudTrail documentation.

Starting today, Amazon RDS for MySQL 26.7 is available in the Amazon RDS Database Preview Environment, allowing you to evaluate the pre-release of MySQL 26.7 on Amazon RDS for MySQL.

MySQL 26.7 is the first Innovation release to adopt the new YY.M (Year.Month) calendar versioning model. It includes bug fixes, security patches, and new features such as Change Stream Applier, a modular replication applier that improves apply throughput and backlog recovery in test workloads. For more details, refer to the MySQL 26.7 release notes.

Amazon RDS Database Preview Environment database instances are retained for a maximum of 60 days and are automatically deleted after the retention period. Database snapshots created in the Preview Environment can only be used within the Preview Environment. Preview Environment database instances are priced according to pricing in the US East (Ohio) Region.

To learn more, visit the Amazon RDS Database Preview Environment documentation.

AWS Continuum for Penetration Testing now available in 6 additional Regions
AWS Continuum for Penetration Testing (AWS Security Agent) is a managed security service that enables AWS customers to conduct  penetration testing against their web applications and APIs. Previously limited in geographic reach, the service now addresses the growing need for localized security testing by expanding into six additional AWS Regions: Asia Pacific (Seoul), Canada (Montreal), Europe (London), US East (Columbus), Europe (Paris), and Europe (Stockholm). This expansion helps organizations operating across these geographies meet data residency requirements while maintaining robust security testing programs.

With this regional expansion, customers can now run penetration testing workloads closer to their production environments. Security and compliance teams can leverage AWS Continuum for Penetration Testing for critical use cases including vulnerability assessments, compliance validation against frameworks such as PCI DSS and ISO 27001, and continuous security posture evaluations.

This expansion is now generally available in Asia Pacific (Seoul), Canada (Montreal), Europe (London), US East (Ohio), Europe (Paris), and Europe (Stockholm), with existing supported Regions remaining fully operational. To learn more about AWS Continuum for Penetration Testing and how to get started in these newly supported Regions, visit https://aws.amazon.com/continuum/.

Amazon RDS for PostgreSQL, MySQL, and MariaDB now support M8a database instances powered by 5th generation AMD EPYC processors, expanding the choice available for your database workloads. Each vCPU in M8a database instances corresponds to a physical CPU core, which delivers consistent per-core performance. For workloads with high I/O requirements, M8a instances provide up to 75 Gbps of network bandwidth and 60 Gbps of Amazon EBS bandwidth. All instances are built on the AWS Nitro System using sixth-generation Nitro Cards.

M8a database instances are available in US East (N. Virginia, Ohio), US West (Oregon), Asia Pacific (Mumbai, Tokyo, Hyderabad) and Europe (Ireland, Frankfurt, Spain) regions. You can get started with M8a database instances through the Amazon RDS Management Console or by using the AWS Command Line Interface (CLI). For the specific engine versions that support these instance classes, see the RDS documentation. For pricing , see the Amazon RDS pricing page.

We're excited to announce in-region support for Anthropic's Claude Opus 5.5 and Claude Sonnet 5 on Amazon Bedrock in the UK (London) Region. Customers who need to process data within the United Kingdom, including those in financial services, healthcare, and the public sector, can now use these models at scale while keeping inference in-country.

In the UK, Amazon Bedrock supports Claude Opus 5.5 and Claude Sonnet 5 with in-region inference on the bedrock-runtime endpoint in the London (eu-west-2) Region. Amazon Bedrock processes inference requests and data within the Region you call, and the processing never leaves that Region.

For the most current information about model availability in each Region, see Regional availability by models in the Amazon Bedrock User Guide. To get started with Claude models on Amazon Bedrock, visit the Amazon Bedrock console.

Amazon Aurora serverless now scales in even larger steps, adding up to 16 ACUs to its current capacity within a second and continuing to scale up to 256 ACUs as your workload grows. With this launch, your workload will scale faster and reach capacity that it requires. When the workload finishes, Aurora serverless automatically scales down to zero, so you only pay for what you use. This makes it especially well-suited for agentic AI applications, which typically have bursts of activity, long idle windows, and unpredictable traffic patterns.

This enhancement is enabled by default on all Aurora serverless clusters running on platform version 3 or 4, with no configuration changes required. Existing clusters on platform versions 1 and 2 can upgrade directly to the latest platform version 4 to benefit from these improvements. You can verify your cluster's platform version in the AWS Management Console under the instance configuration section, or via the RDS API's ServerlessV2PlatformVersion parameter.

For pricing details and Region availability, visit Amazon Aurora Pricing. To learn more, read the Aurora serverless scaling documentation, and get started by creating an Aurora serverless database in just a few steps in the AWS Management Console.

 Today, AWS announces the availability of UltraFast mode for GPT-6 Astra from OpenAI on Amazon Bedrock. Ultrafast is a premium speed tier for GPT-6 Astra built for workloads where speed matters most. According to OpenAI, Ultrafast delivers up to 6x faster inference in the API, with up to 300 tokens per second. The Amazon Bedrock inference engine delivers the performance, security, and reliability required for production workloads.

Use GPT-6 Astra Ultrafast for latency-sensitive applications such as real-time coding assistants, interactive agents, and customer-facing experiences that require fast, high-quality responses. Established AWS controls help you secure workloads, govern access, and audit model invocation activity.

You can get started in the Amazon Bedrock console or programmatically through supported Amazon Bedrock APIs. For information about supported AWS Regions, endpoints, APIs, features, inference profiles and pricing, see the Amazon Bedrock documentation.  

AWS Parallel Computing Service (AWS PCS) now supports scaling logs, which record how AWS PCS scales the compute node groups in your cluster. Each log entry records one state transition for one compute node—for example, an instance launch, a node registration, a scale-down, or a launch failure with its reason. Using scaling logs, you can more easily troubleshoot scaling issues. For example, you can determine why a compute node group did not reach its target size, which launches failed for capacity reasons, and when a specific node started or stopped. Scaling log delivery is opt-in, and you can configure AWS PCS to emit scaling logs to Amazon CloudWatch Logs, Amazon S3, and Amazon Data Firehose.

AWS PCS is a managed service that simplifies running and scaling HPC workloads on AWS using Slurm. You can build complete, elastic environments that integrate compute, storage, networking, and visualization tools, while the service handles cluster operations with managed updates and built-in observability features.

This feature is available in all AWS Regions where AWS PCS is available. To get started, see the AWS PCS User Guide.

Amazon Managed Grafana now supports creating new workspaces with Grafana version 13.2. This release brings features from open-source Grafana versions 13.0 to 13.2, including Git Sync for dashboards, dynamic dashboards, and PromQL query support in the Amazon CloudWatch data source plugin.

With Git Sync, you can treat dashboards as code by linking a workspace to a Git repository, enabling version control for tracking, reviewing, and reverting dashboard modifications. Dynamic dashboards provide a more adaptable approach to layout and paneling, constructing responsive views driven by conditions that respond to data and variables. The Amazon CloudWatch data source plugin now supports PromQL queries, allowing you to use Prometheus Query Language to query metrics that CloudWatch ingests via its OpenTelemetry (OTLP) endpoint, complementing the existing Metric Search and Metrics Insights query types.

Grafana 13.2 is available in all AWS regions where Amazon Managed Grafana is generally available.

To learn more about creating Amazon Managed Grafana workspaces with version 13.2, visit the user documentation. For more information about Amazon Managed Grafana features and pricing, visit the product page and pricing page.

Amazon S3 Vectors now supports pre-filtering, which evaluates metadata filters before running similarity search, returning up to 5x more of the matching vectors when your filter is selective. S3 Vectors also adds a prefix match operator ($startsWith) for filtering on values like paths and URLs. Together, these improvements give your retrieval-augmented generation (RAG), agentic, and semantic-search applications more complete results when you filter, so your applications return more relevant answers.

S3 Vectors provides native support to store and query vectors in Amazon S3, delivering purpose-built, cost-optimized vector storage and query at billion-vector scale. With this launch, indexes in new vector buckets use metadata pre-filtering by default, with no change to how you write vectors with PutVectors or run filtered queries with QueryVectors. To use pre-filtering on an existing index, update it in place with the UpdateIndexMode API. You can also compare pre-filtering against your current filtering on the same index, using a per-query parameter on QueryVectors, before you update.

Metadata pre-filtering is available at no additional cost in all commercial AWS Regions where Amazon S3 Vectors is available, and in the AWS China Regions. We are in the process of deploying this change and plan to complete the deployment in the coming days. To get started, use the AWS CLI, AWS SDKs, or the Amazon S3 console. To learn more, visit the Amazon S3 Vectors documentation and the AWS News blog.

Today, AWS announces system-managed materialized views for Apache Iceberg. Materialized views let you precompute an expensive query once and reuse the result across engines. Because this result is an open Iceberg table in your data lake, anyone with write access can change it, either through the catalog or by writing to the files in Amazon S3. System-managed materialized views close this gap by only allowing AWS Glue to write the materialized view's data and definition, so the result stays exactly as computed regardless of who has write access to the table or the underlying S3 files.

To get started, write the SQL that defines the materialized view and, optionally, a refresh schedule. AWS Glue then computes the results, stores them as a standard Apache Iceberg table in your Amazon S3 Tables bucket, and keeps them current on your schedule. Because the result is an ordinary Iceberg table in the AWS Glue Data Catalog, any Iceberg-compatible engine can read it directly, while the service guarantees no other writer can alter it. You get a governed, always-consistent dataset that you can confidently share. You can still refresh, reschedule, or drop the view whenever you need to.

System-managed Apache Iceberg materialized views are available in all regions where Apache Iceberg materialized views are supported. To learn more, see System-managed materialized views in the AWS Glue Developer Guide.

Amazon WorkSpaces Core Managed Instances now supports Graphics G7 instances, powered by NVIDIA RTX PRO 4500 Blackwell Server Edition GPUs and Intel Xeon 6 processors. G7 instances deliver up to 2.1X better performance for graphics-intensive workloads compared to previous generation G6 instances.

With Graphics G7, customers can run demanding professional applications such as CAD/CAM, 3D rendering, scientific visualization, video editing, and AI-assisted design workflows at higher fidelity and frame rates. G7 instances feature 32 GB of GDDR7 GPU memory per GPU and 2.67X faster memory bandwidth, enabling larger, more complex 3D scenes and models. Six instance sizes are available with 1 to 8 GPUs, vCPUs ranging from 8 to 192, and system memory from 32 GB to 768 GB, with support for Linux and Windows, including bring-your-own-license.

Graphics G7 instances for WorkSpaces Core Managed Instances are available in US East (N. Virginia), US East (Ohio), US West (Oregon), and Europe (Spain). Additional AWS Regions will be added as availability expands.

To get started, specify an instance type from the G7 family when creating a new WorkSpaces Core Managed Instance, or look for G7 instance types when they become available through your WorkSpaces Core partner solution. For more information, see the following resources:

Amazon WorkSpaces Core

Amazon WorkSpaces Core Managed Instances

EC2 G7 Instance Types

Amazon WorkSpaces Core Managed Instances Pricing

Amazon S3 Tables now support up to 100 table buckets per AWS Region in an AWS account, increased from 10. This allows you to create up to 1 million tables per AWS Region in an AWS account. With a higher table bucket allowance, you can create a separate table bucket for each dataset, workload, or team, and apply table bucket-level settings such as encryption, access policies, and replication to each. The higher quota applies by default to all accounts at no additional cost.

S3 Tables deliver the first cloud object store with built-in Apache Iceberg support, and the easiest way to store tabular data at scale. S3 Tables perform continual table maintenance to automatically optimize query efficiency and storage cost over time, even as your data lake scales and evolves.

The higher default quota is available in all AWS Regions where S3 Tables are available. If you need a quota beyond the default, you can request an increase through AWS Support. To learn more, visit the S3 Tables overview page, user guide, and quotas documentation.

​​​AWS Secrets Manager now integrates with the AWS Recommended Actions framework to surface contextual, actionable suggestions for your secrets directly in the Secrets Manager console. ​​

​​With this launch, you can view tailored recommendations alongside your secrets to improve your security posture and follow best practices without leaving the console. For example, you can identify secrets that need rotation configured, use a customer managed key instead of default service-provided key for encryption, and act on secrets configuration improvements, all from a single view.​

​​This feature is available in all AWS Commercial Regions where AWS Secrets Manager is offered, at no additional cost. To get started, visit the AWS Secrets Manager console.​

Amazon EMR Serverless now offers enhanced serverless storage capabilities with support for up to 1TB shuffle operations, raising the previous 200 GB per-job limit. Amazon EMR Serverless makes it simple for data engineers and data scientists to run open-source big data analytics frameworks without configuring, managing, and scaling clusters or servers. This enhancement enables enterprise customers to run production-scale Apache Spark workloads that require processing large volumes of shuffle data during complex operations such as joins, aggregations, and sorting.

Enterprise data teams can now confidently migrate production workloads that routinely process terabyte-scale datasets without worrying about storage constraints. This enhancement is particularly valuable for workloads involving large table joins across multi-terabyte datasets, and complex aggregations on high-cardinality data that require extensive data shuffling. The addition of spill support ensures that jobs can seamlessly handle memory-intensive operations by offloading data to disk when necessary, improving job reliability and success rates for demanding analytical workloads.

This feature is available with Amazon emr-7.14, emr-spark-8.1 and later, in 18 AWS Regions where Amazon EMR Serverless is available. See the Amazon EMR documentation for the full list of supported Regions and their applicable limits.

To learn more about Amazon EMR Serverless and get started with terabyte-scale shuffle support, visit the Amazon EMR Serverless page.

AWS Glue Data Catalog now supports table optimization, statistics, and crawlers for Apache Iceberg Version 3 (V3) tables. With these new capabilities, you can automatically maintain V3 tables, optimize them for query performance, and discover them in Amazon S3.

With table optimization, you can compact V3 tables using binpack, sort, or z-order strategies to improve query performance, and remove expired snapshots and orphan files to reduce storage costs. These optimizations support V3 data types, including variant, geospatial, and nanosecond-precision timestamps. You can also generate number of distinct values (NDV) statistics for V3 tables, which analytics engines use to plan queries efficiently. In addition, you can use Glue crawlers to discover V3 tables stored in Amazon S3 and register them in Glue Data Catalog, making them available to query with any V3-compatible engines.

These capabilities are available for Iceberg V3 tables in all AWS Regions where Glue Data Catalog table optimization, statistics, and crawlers are available. To learn more, see Glue Optimization, Glue Statistics, and Glue crawlers in the Glue Developer Guide.

Route 53 Global Resolver and DNS Firewall now provide DNS analytics and insights through native Amazon CloudWatch integration. These new capabilities enable network administrators and security teams to gain full observability into DNS query patterns, monitor DNS Firewall rule effectiveness, detect anomalous activity, and optimize DNS infrastructure performance. DNS analytics and insights is available in all AWS Regions where Amazon CloudWatch, Route 53 Global Resolver, and DNS Firewall are available.

With CloudWatch Metrics and Contributor Insights, customers can search and analyze DNS query logs, create metric filters for specific patterns such as blocked queries and DNS response codes, and set automated alarms. A new Analytics tab in both the Global Resolver and DNS Firewall consoles provides streamlined access to all analytics in one place.

For example, customers can create a metric filter for blocked DNS queries by VPC and set an alarm to trigger when more than 10 queries are blocked within an hour, enabling rapid response to potential security threats.

Standard Amazon CloudWatch pricing applies to the metrics that customers opt in to. To learn more, visit the Route 53 documentation.

AWS Budgets lets you set custom cost and usage thresholds and alerts you by email when your spending crosses the threshold. AWS Budgets now verifies new email subscribers before sending them budget alerts. When you add an email address to a budget notification, AWS Budgets sends a verification email to that address, and the address begins receiving notifications once the recipient confirms it.

Verification applies to email addresses added from 9/30/26 onward. Addresses already subscribed on existing budgets are unaffected and need no action. For a new address, the recipient confirms it by following the link in the verification email, which opens the AWS Management Console. The AWS Budgets console shows the verification status of every subscriber, with a resend option for addresses still pending confirmation. AWS Budgets delivers these notifications through AWS User Notifications, and recipients can opt out at any time.

AWS Budgets email verification is available in all AWS Regions, except the AWS GovCloud (US) Regions and the China Regions.

To learn more, see Adding email recipients to a budget notification in the AWS Billing and Cost Management User Guide.

Amazon Elastic Kubernetes Service (EKS) Auto Mode now lets you tune kubelet settings, Linux kernel sysctls, and hugepages directly in the NodeClass Kubernetes resource. You can now run workloads that require specific node tuning on EKS Auto Mode, which automates node provisioning, scaling, patching, and upgrades for you.

With the new advancedCompute field in NodeClass, you can configure user namespaces for rootless container builds in CI/CD pipelines, tune eviction thresholds and container log rotation to protect nodes from disk and memory pressure, raise network and ARP cache limits for large clusters, and pre-allocate 2Mi or 1Gi hugepages for HPC, database, and latency-sensitive workloads. You apply these settings declaratively using the Kubernetes API. EKS Auto Mode validates them, applies them at node boot, and preserves them through scaling and upgrade events, so your workloads get the tuning they need while EKS Auto Mode continues to manage the underlying infrastructure, with no EC2 launch templates, custom AMIs, or privileged DaemonSets to maintain.

Amazon EKS Auto Mode advanced compute configuration is available in all AWS Regions where EKS Auto Mode is available.

To get started, see the Amazon EKS Auto Mode product page, and see Create a Node Class for Amazon EKS in the Amazon EKS documentation.

AWS Security Hub now offers remediation plans that group related exposure findings sharing a root cause. Instead of addressing each exposure individually, you can now fix a single underlying resource, such as a misconfigured setting or overly permissive policy, and resolve or reduce the severity of multiple exposures at once.

Each remediation plan includes prioritization guidance (Critical, High, Medium, or Low), impact assessment, and detailed step-by-step instructions with examples in multiple formats including AWS CLI, Terraform, CloudFormation, Python, and CDK. Security Hub automatically prioritizes plans so those reducing the most risk appear first, helping you focus remediation efforts where they matter most. AI agents can also programmatically consume remediation plans through the API to automate security fixes across your environment.

Remediation plans are available in every AWS Region where AWS Security Hub is available and at no additional cost under AWS Security Hub Essentials plan. To learn more, visit the AWS Security Hub documentation.

AWS Transfer Family now lets you choose a custom log group in Amazon CloudWatch Logs for managed workflow execution logs. You can organize workflow logs to monitor individual workflows separately or bring logs from related workflows together.

Previously, every workflow attached to a Transfer Family server sent its execution logs to that server’s CloudWatch log group, so you could not set a separate log destination for each workflow. When creating a workflow through the AWS Transfer Family console or API, you can now select a workflow-level log group that is separate from the server’s log group. Transfer Family delivers workflow logs only to the selected group, so no server logging role is required. Workflow logs retain their existing structured JSON format and remain queryable using Amazon CloudWatch Logs Insights. You can also send logs from multiple workflows to a shared log group to create consolidated metrics and dashboards for tracking workflow execution. If you do not select a structured log destination, workflow logs continue using the server’s role-based logging configuration when configured. Existing workflows continue using their current logging configuration.

This feature is available in all AWS Regions where Transfer Family managed workflows are offered. To learn more, visit the Transfer Family managed workflows User Guide.

Amazon S3 Object Lock support for variable retention with event holds is now available in AWS GovCloud (US-East) and AWS GovCloud (US-West).

Amazon S3 Object Lock variable retention allows you to apply write-once-read-many (WORM) protection to objects whose required retention period starts with a future event, such as a contract closing or an audit completing. You place an event hold with a retention duration on an object and S3 protects the object while the hold is in place. When you release the hold, S3 retains the object for the duration you specified. Unlike legal holds, which end protection immediately upon removal, event holds provide WORM compliance for the required retention period after the triggering event, so you can meet event-based retention requirements without retaining data longer than your policy requires.

To learn more, read the AWS Storage Blog post, the S3 Object Lock overview page, and the S3 documentation. This capability has been assessed by Cohasset Associates for use in environments subject to SEC Rule 17a-4(f), FINRA Rule 4511, and CFTC Regulation 1.31.

On September 30, 2026, Amazon announced a patch update for the following Amazon Corretto Long-Term Support (LTS) version of OpenJDK: Corretto 8u504 is now available for download. Amazon Corretto is a no-cost, multi-platform, production-ready distribution of OpenJDK. This patch includes the tzdata 2026d updates.

Visit Corretto home page to download Corretto 27, Corretto 25, Corretto 21, Corretto 17, Corretto 11, or Corretto 8. You can also get the updates on your Linux system by configuring a Corretto Apt, Yum, or Apk repo.

Feedback is welcomed!

Today, we’re excited to introduce the newest members of the AWS Heroes program. AWS Heroes are a vibrant, worldwide group of AWS experts who go above and beyond to share knowledge, mentor others, and build thriving communities. These individuals make a real difference in helping developers and organizations succeed with AWS. This month, we welcome […]

Amazon Aurora PostgreSQL now lets you directly query Apache Iceberg and Parquet data stored in your data lake alongside live operational data—no ETL pipelines required. Powered by DuckDB embedded within Aurora, this capability enables single queries that join transactional and historical data using familiar PostgreSQL syntax. It supports AWS Glue Data Catalog, S3, and S3 Tables, with optimizations like predicate pushdown and caching for efficient performance.

Amazon S3 Vectors now supports metadata pre filtering, delivering up to 5x higher recall on filtered searches. Filters evaluate before similarity search so scoped queries return more relevant results. Ideal for RAG, agentic apps, and document search.

Amazon S3 Tables now offer complete support for the Apache Iceberg V3 spec. Upgrade existing V2 tables or create new V3 tables with built-in compaction, maintenance, and engine compatibility.

オリックス株式会社の SaaS「PATPOST」の既存 REST API を、Amazon Bedrock AgentCore Gateway で MCP サーバー化した取り組みをご紹介します。既存資産に手を入れずに MCP サーバー化する方法を解説します。

Amazon Aurora PostgreSQL で、データレイクに保存された Apache Iceberg および Apache Parquet 形式のデータを直接クエリできるようになりました。DuckDB が Aurora PostgreSQL に組み込まれ、ETL パイプラインなしで運用データとデータレイクのデータを 1 つのクエリで横断的に分析できます。

本日、Kiro workflows を発表します。複数のエージェントを使い、少ない監督で複雑なタスクを最初から最後までやり遂げられる機能です。私たちは Kiro 自体の開発にも workflows を使ってきました。新しいクラウド構成、クラウドセッション、そして workflows 体験の大部分がその成果です。

Amazon Bedrock Guardrails の自動推論チェックを活用し、大規模言語モデル (LLM) の回答を反復的に書き換えて改善するオープンソースのサンプルチャットボットを紹介します。検出結果の種類 (VALID、INVALID、SATISFIABLE など) に応じた優先順位付けと書き換えループの仕組み、確認のための質問を行うロジック、数学的に検証可能な説明と監査証跡による透明性の向上、バックエンドの主要コンポーネントと導入手順を詳しく説明します。

Amazon Aurora および Amazon RDS（MySQL / PostgreSQL）と Amazon OpenSearch Service の統合が一般提供開始されました。Amazon OpenSearch Ingestion を使うと、ETL パイプラインを自前で組まずに、データベースの内容をニアリアルタイムで OpenSearch に同期できます。本記事では、サンプルの HR データベースを使って設定の流れと同期の動きを紹介します。

2026年7月29日から 2 日間にわたって、アマゾンウェブサービスジャパン合同会社はパシフィコ横浜で開催された KubeCon + CloudNativeCon Japan 2026 においてスポンサーとしてブースを出展しており、たくさんの方にご来場いただきました。（ご来場いただいた皆様、有難うございました！）  AWS ブースでは、2日間で 12 のデモセッションを行い、多くのご来場者にご視聴いただきました。その際にセッション資料をシェアしてほしいとのご要望をいくつか頂いたことを受け、こちらのブログで各セッションの資料を公開させて頂きます。

世界最大の株式デリバティブ清算機関である The Options Clearing Corporation (OCC)® は、Amazon Bedrock で構築した AI 搭載のセキュリティ調査エージェント「SOC エージェント」を導入しました。本記事では、エージェンティックループによってセキュリティアラートの調査を自動化し、解釈可能かつ監査可能な結果を提供する仕組みを紹介します。また、ヒューマンインザループのフィードバックによって、セキュリティ専門家が主導権を保ちながらエージェンティック AI を活用する方法も説明します。

2026 年 9 月 21 日週を表現するテーマが 1 つあるとすれば、「選択」ではないでしょうか。フロンティ […]

AWS の AI 搭載エンジニアリングエージェント Kiro をベースにしたサンプルエージェントを公開しました。カスタム SAP ABAP プログラム、SAP BW、SAP PI/PO のモダナイゼーションを単一のプロンプトで一括実行でき、社内ベンチマークでは ABAP コードのモダナイゼーション工数を最大87%削減しています。SAP S/4HANA 準拠への変換、クリーンコア化、SAP BTP Integration Suite や SAP Datasphere への移行、仕様書生成、単体テスト自動化といったユースケースと、正確性を担保するセーフガード、始め方を紹介します。

Kiro ベースのサンプルエージェントを使った SAP モダナイゼーションの6つのユースケースについて、社内ベンチマークとパートナー・お客様の成果を詳しく紹介します。SAP ECC の ABAP コード88オブジェクトを4.5時間で SAP S/4HANA 準拠に変換して工数を87%削減したほか、SAP PI/PO の22インターフェースを約2時間で SAP BTP Integration Suite へ、SAP BW 環境を2.2時間で SAP Datasphere へ変換しました。依存コードの取得、ステアリングファイル、単体テスト、意思決定ログといった、正確性と監査可能性を支えるセーフガードも解説します。

MHK built the SmartProminence AI Orchestrator, a HIPAA-eligible agentic workflow framework on AWS using Amazon Bedrock, Amazon ECS, and event-driven patterns. It reduced manual medical review effort by 90% and cut new AI feature deployment from months to weeks. Learn how MHK architected this reusable, multi-tenant orchestrator for healthcare AI.

Prove your multi-AZ architecture can sustain a real impairment. This post shows how to run a multi-day (48-72 hour) Availability Zone evacuation drill with ARC Zonal Shift across Amazon ECS, Amazon EKS, Amazon RDS for PostgreSQL, and Amazon Aurora PostgreSQL, with step-by-step CLI commands, prerequisites, observability metrics, and restore procedures.

Amazon Bedrock AgentCore Runtime Instances gives multi-agent workflows AWS managed EC2 infrastructure with GPUs, persistent volumes, and multi-day sessions. In this post, we deploy a three-agent music production pipeline where the agents colocate on one GPU instance, share a filesystem, and hand work to each other to produce a finished track.

This technical how-to builds a conversational claims assistant on Amazon Bedrock Knowledge Bases that answers natural-language questions with citations. It covers ingesting claim documents from Amazon S3, querying with the AgenticRetrieveStream API, multi-turn follow-ups, metadata filters, and contextual grounding guardrails.