All news in category “Vendor and Hyperscaler Watch”

📍Amazon DocumentDB (with MongoDB compatibility) is now available in the Asia Pacific (Melbourne) region. Amazon DocumentDB is a fully managed, native JSON document database that automatically scales storage up to 128TiB and supports millions of requests per second. It can be scaled out to 15 low-latency read replicas in minutes with no application downtime and integrates with services such as AWS DMS, CloudWatch, CloudTrail, Lambda, and Backup. Clusters can be created via the AWS Management Console, CLI, or SDKs.

🚀 Amazon Web Services announced that Amazon DocumentDB (with MongoDB compatibility) is now available in the Europe (Zurich) region. The fully managed, JSON-native document database provides automatic storage scaling up to 128TiB and can support millions of requests per second while scaling out to 15 low-latency read replicas with no application downtime. It integrates with AWS DMS, CloudWatch, CloudTrail, Lambda and AWS Backup, and clusters can be created via Console, CLI, or SDK. Local availability improves latency and data residency options for Swiss and nearby customers.

🎧 Amazon Connect launches Audio Enhancement to improve voice clarity and reliability by suppressing agent-side background noises and isolating agent voices in busy contact centers. The feature includes two modes: Voice Isolation (removes background speech and noise) and Noise Suppression (targets non-speech noise). Administrators can enable modes via the User Management page, and agents with permissions can adjust settings in the Contact Control Panel. This capability is available in all commercial AWS Regions where Amazon Connect is offered.

🔧 This guide explains how to adapt foundation models on Google Cloud by fine-tuning both managed and self-managed workflows. It contrasts a fully managed Vertex AI Supervised Fine-Tuning path for models like Gemini with a customizable GKE approach using LoRA on open-source models such as Llama. The labs walk through data preparation, baseline evaluation, tuning, and automated evaluation metrics, as well as GKE infrastructure, GPU provisioning, security with Workload Identity, and containerized training for production readiness.

🧰 The Google Cloud samples team describes building a specialized end-to-end system that uses Gemini on Vertex AI and Genkit to produce production-ready educational code samples across many languages and products. Their architecture separates generation, validation, and delivery so LLM outputs are combined with deterministic automations, linters, unit tests, and human review. The post presents seven practical technical takeaways—decomposition, determinism, precise prompts, vetted evaluation, scaled downstream processes, end-to-end testing, and solid engineering practices—that drove reliable, scalable sample generation.

🔒 Amazon Bedrock AgentCore Browser now accepts customer-provided proxy configurations, allowing organizations to route browser sessions through corporate or regional proxy infrastructure for geo-targeting, compliance, and stable egress addresses. The feature supports both HTTP and HTTPS protocols and integrates with AWS Secrets Manager for secure credential management. It is available in all 14 regions where AgentCore Browser is offered.

⚙️ Falcon Fusion SOAR simplifies SOC automation by enabling teams to start with single, high-impact workflows and scale to agentic, AI-driven orchestration. New capabilities — natural language Workflow Generation, a Test-and-Debug preview, and a Data Transformation Agent powered by Charlotte AI — lower the barrier to building reliable automations. It integrates endpoint, identity, cloud, and threat intelligence, keeps humans in the loop, and supports mature programs that adopt Charlotte Agentic SOAR for agent orchestration.

🛡️ Breach & Attack Simulation (BAS) tools automate validation of security controls by emulating adversary tactics and mapping those behaviors to frameworks such as MITRE ATT&CK or the Cyber Kill Chain. BAS focuses on verifying detection and prevention capabilities across endpoints, networks, mail gateways and identity systems rather than discovering unknown vulnerabilities. When evaluating products, prioritize realistic, customizable scenarios, scalable automated testing, clear reporting, integration with existing tooling, and vendor support or managed options.

🖥️ Microsoft has released Windows 11, version 26H1, but the distribution is limited to new ARM-based devices such as those using Qualcomm Snapdragon X2 processors and possibly other upcoming ARM chips. This build is a platform-specific release and will not be broadly delivered through Windows Update to existing PCs. Microsoft says organizations should continue to purchase, deploy, and manage devices on broadly released versions like 24H2 and 25H2. The company also reiterated its annual update cadence, with 26H2 expected later in the year.

🔁 AWS now supports managed minor version upgrades for Aurora Global Database, enabling you to upgrade an entire global topology with minimal downtime. Administrators can initiate upgrades from the AWS Management Console, SDK, or CLI and have all regional clusters automatically moved to the chosen minor version. This eliminates manual per-cluster upgrades and reduces operational overhead for global cluster management. The capability currently supports Aurora PostgreSQL-compatible engines and is available in all commercial AWS Regions and AWS GovCloud (US) Regions.

🔧 Amazon SageMaker HyperPod now lets operators manage individual cluster nodes directly from the AWS Console. The console enables SSM session launches, copyable pre-populated SSM CLI commands, and direct node actions such as reboot, delete, and replace, with support for batch operations across multiple nodes. Available in all Regions where HyperPod is supported, these controls reduce context switching and speed manual recovery for time-sensitive AI training and inference workloads.

🚀 Amazon Web Services expanded availability of EC2 C8id, M8id, and R8id instances powered by custom Intel Xeon 6 processors, offering up to 384 vCPUs, 3 TiB of memory, and 22.8 TB of NVMe SSD storage. These instances deliver up to 43% higher compute performance and 3.3x more memory bandwidth versus the sixth-generation C6id, M6id, and R6id, and provide up to 46% better I/O performance for database workloads. An Instance Bandwidth Configuration lets customers flexibly allocate 25% between network and EBS; C8id and M8id are now in Europe (Frankfurt) and Asia Pacific (Tokyo), with M8id and R8id additionally in Europe (Spain), and the family is purchasable via On‑Demand, Savings Plans, and Spot.

🔒 Google Distributed Cloud (GDC) air-gapped 1.15 introduces networking updates that give regulated environments more control, visibility, and scalability while preserving isolation. Preview features include Cloud NAT for configurable egress IPs and timeouts, enhanced connectivity enabling standard clusters to reach organizational workloads securely, and HTTP/HTTPS load balancer health checks for application-level monitoring. IPAM's subnet group capability (GA) simplifies subnet scaling by letting child subnets reference multiple parent subnets.

🚀 Google has opened the Gemini Enterprise Agent Ready (GEAR) learning program to all developers and professionals as a dedicated pathway within the Google Developer Program. GEAR provides 35 monthly Google Skills credits for hands-on labs and sandbox testing and features end-to-end learning paths such as Introduction to Agents and Develop Agents with the ADK. Members can earn profile badges and pursue instructor-led certification tracks to validate enterprise agent engineering skills and accelerate production-ready deployments.

⚡ Amazon Athena now offers 1-minute Capacity Reservations and lowers the minimum reserved capacity to 4 Data Processing Units (DPUs). The feature provides dedicated serverless compute for workloads requiring query prioritization and concurrency controls, with no long-term commitments. You pay only for the capacity you reserve and there are no data-scanned charges; reservations attach to existing workgroups without SQL or application changes. AWS cites up to 95% cost savings for short-duration query workloads and enables more frequent, fine-grained capacity adjustments to match workload patterns.

🧭 Amazon Bedrock now supports six open-weights models — DeepSeek V3.2, MiniMax M2.1, GLM 4.7, GLM 4.7 Flash, Kimi K2.5, and Qwen3 Coder Next. These models span frontier reasoning, agentic intelligence, and autonomous coding while offering lower-cost inference options for enterprise workloads. They run on Project Mantle, a distributed inference engine that delivers serverless, high-performance model serving with OpenAI API compatibility, automated capacity management, quality-of-service controls, and higher default quotas for production deployment.

📥 Amazon Elastic Kubernetes Service (EKS) Auto Mode can now deliver logs via Amazon CloudWatch Vended Logs. Customers can configure each managed capability—compute autoscaling, block storage, load balancing, and pod networking—as a vended log source using the CloudWatch APIs or the AWS Console. Logs can be routed to CloudWatch Logs, Amazon S3, or Amazon Kinesis Data Firehose. This option uses built‑in AWS authentication and authorization and can reduce delivery cost versus standard CloudWatch Logs.

⚡Microsoft is evaluating high-temperature superconductors (HTS) to improve datacenter power delivery, reduce transmission losses, and increase capacity within compact footprints. HTS cables, when maintained at cryogenic temperatures, carry electricity with near-zero resistance, enabling smaller, lighter cabling that generates negligible heat and avoids voltage drops. By partnering with manufacturers and system integrators, Microsoft plans pilots of rack-level HTS architectures and aims to rethink traditional power designs to support AI-era workloads.

🔐 Amazon OpenSearch Serverless now supports Collection Groups, enabling you to share OpenSearch Compute Units (OCUs) across collections that use different AWS KMS keys. This shared compute model reduces overall OCU costs while preserving collection-level security and access controls. Collection Groups also let you set minimum and maximum OCU allocations, allowing predictable startup capacity and eliminating cold-start latency for latency-sensitive workloads.

⚠️ Microsoft is investigating an outage that prevents some administrators with business or enterprise subscriptions from accessing the Microsoft 365 admin center in North America and Canada. The company is tracking the issue on its service health page and is collecting telemetry, with an early focus on CPU utilization and user HTTP Archive (HAR) files to identify a root cause. Impacted users report slow or unavailable admin portal access, degraded functionality, and potential inability to open the M365 app or raise support tickets.