Category: Compute

Open-weight large language models are being released by research institutions worldwide, but turning published weights into production inference services remains a challenge—especially under strict data residency requirements. This post shows how Public AI built a scalable inference platform on Amazon EKS and Intel-powered Amazon EC2 instances to serve Switzerland’s Apertus model family, and why this architecture provides a repeatable blueprint for sovereign LLM initiatives.

Learn how Fovus, an AWS Advanced Technology Partner, eliminates the inefficiency of static Nextflow resource configurations by benchmarking each pipeline process individually and applying AI-powered orchestration to select optimal EC2 instance types, memory, storage, and parallel computing settings—delivering up to 85% cost savings on genomics workloads running on AWS.

SAS Viya Workbench on AWS provides a unified cloud environment for Python, R, and SAS, letting teams focus on building models instead of managing infrastructure. Its split-plane architecture keeps your data securely within your AWS account while SAS handles orchestration. With elastic scalability, pre-configured environments, and built-in collaboration, teams can accelerate analytics innovation from day one.

Nutanix Government Cloud Clusters (GC2) on AWS GovCloud (US) enables federal agencies and Department of War organizations to accelerate cloud migration while maintaining strict security and compliance standards including FedRAMP High, CJIS, ITAR, and DoW Impact Levels 4/5. This solution bridges the gap between cloud innovation and mission-critical security requirements, allowing agencies to migrate legacy workloads without refactoring while using familiar Nutanix tools across hybrid environments.

AVEVA and AWS are strengthening their collaboration to help industrial organizations reduce IT challenges and accelerate cloud adoption. Building on their 10-year relationship, AVEVA has successfully validated key components of its industrial operations software portfolio on AWS infrastructure, including AVEVA System Platform and Manufacturing Execution System on Amazon EC2, plus integration of Amazon RDS with AVEVA PI Data Infrastructure. Organizations can now manage their entire operational technology stack within a unified AWS environment, addressing critical pain points like system maintenance, data integration challenges, and total cost of ownership.

OPSWAT’s MetaDefender Storage Security delivers automated file security for regulated financial institutions on AWS, combining Deep Content Disarm and Reconstruction with multi-scanning and data loss prevention capabilities. A major U.S. bank achieved 75% faster security update deployments while maintaining compliance with GDPR, HIPAA, and GLBA through containerized deployment on Amazon ECS and EKS with seamless S3 integration. Available through AWS Marketplace, the solution provides continuous monitoring, automated compliance reporting, and reduced operational costs for organizations managing sensitive data across multiple cloud regions.

Queue-it, an AWS Partner, integrates with Amazon CloudFront to provide virtual waiting room capabilities that control visitor flow during high-traffic events—addressing scenarios where traffic spikes faster than infrastructure can scale, backend systems hit constraints, or fairness matters for limited-inventory sales. Built entirely on AWS services including DynamoDB, Lambda@Edge, and Amazon EC2, Queue-it’s solution helped Sky Mobile achieve a 37% conversion rate increase during their iPhone launch and enabled His Majesty’s Passport Office to handle traffic spikes of nearly 70,000 applications while maintaining system stability.

Securonix, a six-time Gartner Magic Quadrant Leader for SIEM, processes more than 78 million security events per second to help organizations reduce cybersecurity threats in near real time. Facing rising data volumes and an increasingly complex threat landscape, the company partnered with AWS Enterprise Support to modernize its infrastructure, migrating Spark processing workloads to Amazon EKS across 40+ clusters executing over 10,000 concurrent jobs. The transformation delivered impressive results: 30% improvement in cost performance, 99.99% availability, 91% faster code deployment (from 3-4 hours to 15 minutes), and 50% reduction in job recovery time. This modernization not only freed engineering teams from maintenance overhead but also established a scalable foundation for future growth, including planned integration of Amazon Bedrock’s generative AI capabilities.

Discover how Blux, a B2B SaaS startup, successfully implemented their personalized product recommendation service using AWS EKS architecture. Follow their technical journey of building an AI-powered recommendation system that delivers 7x improvement in conversion rates with just 2-3 hours of implementation time.

Discover how Bitrise leverages Amazon EC2 Mac instances to streamline and accelerate iOS build automation. This post explores how developers can run scalable, secure CI/CD pipelines in the cloud to reduce build times and improve delivery efficiency for mobile apps.