All news in category “Vendor and Hyperscaler Watch”

🚀 Amazon Managed Service for Apache Flink now supports Apache Flink 2.2, offering runtime improvements including Java 17 support, RocksDB 8.10.0 for improved I/O, and serialization enhancements. The release deprecates the Dataset API and Scala APIs. You can create new Flink 2.2 applications or perform in-place version upgrades across AWS regions to simplify adoption. These updates aim to improve performance, reliability, and upgrade velocity for streaming workloads.

🔍 Amazon SageMaker Unified Studio now surfaces Amazon CloudWatch metrics for AWS Glue jobs directly alongside job logs in a single, unified interface. Data engineers can correlate DPU utilization, memory consumption, CPU load, and data movement size with log output to diagnose compute bottlenecks and memory pressure faster. The consolidated view reduces mean time to resolution for ETL pipelines and is available in all Regions where SageMaker Unified Studio is generally available. To view metrics, open a Glue job run and select the Metrics tab.

🧭 AWS Organizations API responses now include the complete organizational path for accounts and organizational units, eliminating the need to traverse hierarchies with multiple calls. APIs such as DescribeAccount, ListAccounts and DescribeOrganizationalUnit return the full path (for example: o-{orgId}/r-{rootId}/ou-{ouId}/{accountId}) in a single response. This reduces operational overhead, lowers latency and simplifies troubleshooting, governance and automation workflows. The capability is available in all commercial Regions and AWS GovCloud (US) Regions, helping enterprises and regulated customers gain immediate org context.

☁️ AWS Deadline Cloud introduces three new fleet scaling options—worker idle duration, standby worker count, and scale out rate—to give creative teams finer control over render farm capacity and responsiveness. Worker idle duration keeps instances available after jobs complete to reduce cold starts and speed artist iteration. Standby worker count maintains a pool of pre-warmed workers for immediate job starts, while scale out rate lets you control expansion speed up to 500 workers per minute to match infrastructure and budget needs.

🔒 Google has begun rolling out Android developer verification, requiring developers who distribute apps outside Google Play to create an account in the Android Developer Console to confirm their identity. The rollout precedes a September enforcement in Brazil, Indonesia, Singapore, and Thailand, with global expansion planned next year. Sideloading of unregistered APKs remains possible for power users via an advanced flow that includes an authentication step and a one-off 24-hour waiting period to deter scammers.

🔁 AWS Marketplace now offers sellers a streamlined self-service workflow for refunds and agreement cancellations that removes the need to file support tickets and provides full visibility into request status. Sellers can create requests from the Agreements page or programmatically via the AWS Marketplace Agreement APIs; requests are pre-populated with agreement and invoice data and processed automatically. Buyers can approve cancellations in-console and see refunds reflected on their charge summary, and KYC verification is limited only to invoices that require compliance validation.

🔐 The new Aurora DSQL connectors for .NET (Npgsql) and Rust (SQLx) simplify secure application access by automating IAM token generation, SSL setup, and connection pooling. They remove reliance on static user passwords while remaining fully compatible with existing driver features. The connectors also provide opt-in optimistic concurrency control retries with exponential backoff, custom IAM credential providers, and AWS profile support to ease credential management.

🚀 Amazon expanded S3 Vectors into 17 additional AWS Regions — now available in 31 Regions worldwide. S3 Vectors is the first cloud object storage with native vector support, built for AI agents, inference, Retrieval-Augmented Generation (RAG), and semantic search at billion-vector scale. It supports up to two billion vectors per index, elastic scaling to 10,000 vector indexes per bucket, low-latency queries (frequent queries as fast as 100 ms; infrequent under one second), and native integration with Amazon Bedrock Knowledge Bases to help reduce RAG costs.

📩 AWS announced AWS End User Messaging Notify, a service that lets developers send one-time passcodes (OTPs) within minutes using phone numbers and sender IDs owned by AWS. Developers configure a brand name, enable SMS, voice, or both, and use ready-to-use templates to send messages to over 200 countries. Every API call includes built-in SMS fraud protection via AWS End User Messaging SMS Protect at no extra cost, and spend limits can pause delivery if thresholds are met. Notify is available in all AWS Regions where End User Messaging is offered.

⚠️ AWS announced availability changes across multiple products: some services and features are moving to Maintenance, a set of services are entering Sunset, and one feature has reached End of Support. Starting April 30, 2026, services designated as Maintenance will no longer be accessible to new customers, while existing customers may continue use. Notable items include Amazon ARC Readiness Check, Amazon Comprehend features, AWS App Runner, and Amazon RDS Custom for Oracle. AWS is providing migration guides and support to assist affected customers.

🚀 Google Cloud Spanner’s interoperable multi-model capabilities are being adopted by customers to run relational, graph, vector, and full-text workloads together at global scale. Real-world examples across fraud detection, recommendation engines, hybrid search, and autonomous network operations show consolidation of database sprawl, elimination of ETL, and improved real-time analytics. Customers such as DANA, Palo Alto Networks, Target and Inspira highlight Spanner’s global consistency, native graph and vector search, and high availability as enablers of simpler, future-proof architectures.

🔍Spanner positions itself as a unified, globally consistent database designed for agentic AI by combining relational, key-value, graph, vector and full-text search capabilities in one platform. The post argues this interoperable multi-model approach reduces data silos, removes brittle synchronization logic, and improves governance, availability, and development velocity. Google highlights features such as GQL graph support, a Cassandra native endpoint for lift-and-shift, and ScaNN-based ANN vector search. The customer example of MakeMyTrip illustrates significant operational simplification and faster AI feature delivery.

🛡️ In 2025 Google’s Vulnerability Reward Program (VRP) celebrated its 15th anniversary and awarded over $17 million to more than 700 researchers worldwide — a 40%+ increase versus 2024. The year introduced a standalone AI VRP, extended Chrome rewards for AI features, and launched a patch rewards program for OSV-SCALIBR. Multiple bugSWAT events and the ESCAL8 conference generated hundreds of reports and significant payouts. Google reaffirms its commitment to collaboration, transparency, and continued events in 2026.

💽 Amazon ECS Managed Instances now supports using Amazon EC2 instance store volumes as a data volume option for container workloads. You can enable local storage by configuring a custom ECS Managed Instances capacity provider and selecting EC2 instance types that include instance store volumes; when an instance lacks instance store or local storage is disabled, Amazon ECS will automatically provision an Amazon EBS data volume. This reduces storage costs and can accelerate I/O for latency-sensitive workloads. Support is available in all commercial AWS Regions where ECS Managed Instances is offered.

🔐 AWS Security Agent is now generally available, offering on-demand, continuous penetration testing across AWS, Azure, GCP, other cloud providers, and on-premises environments. The service deploys autonomous AI agents that combine SAST, DAST, and active exploit attempts to validate findings, reduce false positives, and provide CVSS-scored, reproducible results. Pricing is metered at $50 per task-hour, the product supports authenticated flows via LLM-driven sign-ins, and includes automated remediation suggestions and pull requests to accelerate fixes.

🎯 Amazon Bedrock AgentCore Evaluations is now generally available to deliver automated, continuous and on-demand quality assessment for AI agents. The feature provides online evaluation to sample and score live production traces and on-demand evaluation for programmatic tests in CI/CD pipelines and interactive workflows. It includes 13 built-in evaluators covering response quality, safety, task completion and tool usage, plus Ground Truth and customizable LLM- or code-based evaluators.

⚡Active Buffer is a GKE preview that implements the Kubernetes CapacityBuffer API to remove scale-out latency by keeping spare node capacity warm. It replaces manual 'balloon' pod hacks and costly over-provisioning with a declarative resource the Cluster Autoscaler treats as pending demand, so critical pods can land instantly. Buffers can be sized by fixed replicas, percentage of deployments, or resource limits.

🔒 Microsoft and Armada are integrating Azure Local and Foundry Local into Armada’s Galleon modular datacenters to provide a customer-controlled cloud for intermittently connected, contested, or fully disconnected environments. The validated reference architecture supports Azure Local control plane and managed clusters with multi-rack scalability, flexible hyperconverged or SAN-backed storage, and resilient multi-network connectivity including satellite, LTE/5G, RF, and SD-WAN. Designed for defense, public safety, energy, and other regulated sectors, the solution preserves Azure’s cloud operating model while enabling local AI inference and analytics to meet sovereignty, latency, and regulatory requirements.

⚙️ Google and partners are applying generative AI to streamline environmental reporting and infrastructure. Internally, Google used Gemini to auto-validate claims against policies and NotebookLM to convert static reports into an interactive, cited knowledge base. Equinix built a Sustainability Data Lake on BigQuery, ingesting data from 240+ sites to shift from manual spreadsheets to on-demand insights. The approach pairs serverless architecture and carbon-intelligent infrastructure to cut cost, compute waste, and reporting cycle time.

📱 AWS End User Messaging now supports Rich Communication Services (RCS) for Business, enabling brands to deliver verified, branded messages that display company name and logo within native messaging apps. The service lets you create and manage RCS agents in the console or via APIs and includes automated SMS fallback when recipients lack RCS support. Integration uses existing AWS event and messaging pipelines to enable bidirectional, AI-driven conversational experiences.