Lumen Intelligent Internet bridges owned global fiber scale with software-defined agility, delivering on-demand 100 Gbps elasticity, inline security, and pay-as-you-grow economics to eliminate over-provisioning for AI and cloud workloads.

OpenSearch is expanding from enterprise search into retrieval, memory, and observability for AI applications, but the market still needs a better way to translate business questions into reliable searches.

Perplexity Portable on AMD Ryzen AI Max signals a shift to local AI workloads, enhanced data privacy, and local-first multi-step agentic automation.

Microsoft Foundry accelerates production-grade AI scaling by unifying model-agnostic orchestration, token-efficient state preservation, and enterprise governance to continuously optimize multi-agent workflows.

Broadcom VCF enables European enterprises to bridge the hybrid strategy execution gap by providing a consistent operating model that pairs with sovereign infrastructure partner OVHcloud, driving compliant cloud modernization without application re-architecture.

Row Zero will add a live spreadsheet interface to Genie, giving business users a familiar way to analyze governed enterprise data without relying on static exports.

NVIDIA adds software and infrastructure controls designed to keep agents within authorized boundaries when application safeguards fail.

The new architecture gives agents access to current operational data without making their queries compete with production database workloads.

From ZT Systems to Silo AI to World Labs, AMD keeps moving up the stack, now to the research that defines what the stack must run

Confidential execution extends the VAST AI Operating System into model placement, protecting enterprise data and proprietary model weights while both are in use.