AWS Media Services is a comprehensive portfolio of purpose-built, managed cloud services for video transport, preparation, processing, and delivery. They enable customers to build and adapt live and on-demand video workflows quickly without capacity planning, scaling seamlessly with pay-as-you-go pricing. Built on AWS's secure global infrastructure, these services deliver industry-leading video quality and resiliency, powering the world's largest streaming events. Their integration with other AWS services and third-party applications for storage, machine learning, content protection, and monetization provides a differentiated, end-to-end cloud solution for media and entertainment.
Set up a render farm in minutes
Scalable and flexible compute render management
Render production-grade volumetrics
Build your particle meshes from multiple sources
Simplify particle simulations for Autodesk 3ds Max
Increase file loading speeds for animated scene geometry
Reliable, secure, and flexible transport for live video
Prepare file-based video assets for on-demand broadcast and multiscreen delivery
Convert inputs into live outputs for broadcast and multiscreen video delivery
Originate and package live and on-demand video content
Originate and store video assets for live or on-demand media workflows
Personalize and monetize multiscreen video content with server-side ad insertion
Build engaging live stream experiences with a managed video service
Capture, process, and store video streams for analytics and machine learning
Convert inputs into live outputs for broadcast and multiscreen video delivery
Manage video networks for live and on-demand content
Process live video channels using statistical multiplexing
Connect a live video source to AWS
Matt Garman
Chief Executive Officer, Amazon Web Services
AWS Lambda has launched a public preview for managed runtimes supporting Node.js 26 and Python 3.15. This initiative allows developers and third-party tool providers to test compatibility and provide feedback before the runtimes reach general availability.
AWS has launched the AWS Agent Registry in preview and contributed to the new Agentic Resource Discovery (ARD) open specification. These tools aim to provide a standardized, searchable catalog for AI agents and Model Context Protocol (MCP) servers to facilitate cross-environment resource discovery.
AWS has released the Agentic Data Operations Platform (ADOP), a reference architecture that utilizes Amazon Bedrock and Claude to automate data engineering tasks like ETL, quality checks, and semantic modeling. The platform is designed to accelerate data source onboarding by generating deterministic code artifacts while enforcing architectural and compliance guardrails.
Amazon EKS has introduced automated certificate authority (CA) rotation to help manage cluster security and prevent expiration for AWS-managed components. While AWS automates the lifecycle for managed nodes, customers remain responsible for updating worker nodes and external clients to maintain secure API connections.
AWS has introduced inbound prefix controls for its Direct Connect service, increasing the maximum route prefix limit from 100 to 1,000 for private and transit virtual interfaces. This update allows streaming infrastructure teams to manage prefix capacity more granularly across connection and gateway levels.
Amazon Web Services has expanded the regional availability of its Graviton4-powered EC2 C8gd, M8gd, and R8gd instances. These instances provide high-speed local NVMe storage and improved performance for compute-intensive streaming workloads such as video encoding and real-time analytics.
Amazon Redshift has introduced long-term system table retention by enabling automatic data replication to Amazon S3 Tables in Apache Iceberg format. This update allows engineers to bypass custom ETL pipelines for cross-warehouse observability and performance analysis.
Amazon has introduced a visual interface for Generative AI Inference Recommendations within its SageMaker AI Studio. The tool utilizes NVIDIA AIPerf to benchmark model configurations, helping engineering teams optimize production workloads for latency, throughput, and cost.
AWS has detailed its strategy for integrating vector search capabilities directly into six of its existing data services, including OpenSearch, S3, and DynamoDB. These native integrations aim to support agentic AI workflows, such as RAG and multimodal content discovery, without requiring customers to migrate data to specialized vector databases.
AWS has published architectural guidance for scaling agentic AI systems within enterprise environments using Amazon SageMaker and Amazon Bedrock. The framework focuses on separating control and execution planes to manage multi-model and multi-framework environments while avoiding vendor lock-in.
