Platform Engineering

Microsoft AKS at KubeCon 2026: GPU Scheduling and Fleet Control

Microsoft AKS at KubeCon 2026: GPU Scheduling and Fleet Control

Microsoft announced major AKS and Fleet Manager updates at KubeCon + CloudNativeCon North America 2026, covering managed Dynamic Resource Allocation for GPUs, multi-cluster placement controls, upgrade safety tooling, and hardware-backed security for regulated workloads. The announcements position Kubernetes as the scheduling layer for heterogeneous AI infrastructure. Platform teams running fleets at scale will find the most immediate operational value in drift detection, upgrade health gates, and the confidential VM node pool support.

Microsoft AKS at KubeCon 2026: GPU Scheduling and Fleet Control Read More »

VictoriaStack: One Open-Source Observability Stack for Metrics, Logs & Traces

VictoriaStack: One Open-Source Observability Stack for Metrics, Logs & Traces

VictoriaMetrics announced VictoriaStack, an open-source observability platform that brings metrics, logs, and traces together under a single Kubernetes Operator and OTLP-native architecture. The stack includes MCP server integrations for AI agent access, positioning it squarely for agentic era infrastructure demands. ECI Research data shows developer velocity is the top technical selection factor, making VictoriaStack’s reduced operational complexity a genuine competitive differentiator.

VictoriaStack: One Open-Source Observability Stack for Metrics, Logs & Traces Read More »

Cloudsmith Launches AI Agent Skills Registry at KubeCon 2026

Cloudsmith Launches AI Agent Skills Registry at KubeCon 2026

Cloudsmith announced its AI Agent Skills Registry at KubeCon 2026, extending enterprise artifact governance to the AI coding agent skills that autonomous agents execute during software development. The move addresses a growing software supply chain gap as agentic workflows expand beyond traditional developer populations. Platform engineering teams gain version control, policy enforcement, and audit observability for AI skills alongside existing artifact types.

Cloudsmith Launches AI Agent Skills Registry at KubeCon 2026 Read More »

Broadcom Expands Kubernetes Platform with VKS 3.8 and CNCF Contribution

Broadcom Expands Kubernetes Platform with VKS 3.8 and CNCF Contribution

Broadcom has released VMware vSphere Kubernetes Service 3.8 and contributed VM API to the CNCF Sandbox, targeting enterprise Kubernetes operational friction with automated upgrades, OPA-backed compliance bundles, and native developer tooling. The validated partner ecosystem now spans observability, security, AI/data, and storage vendors. ECI Research data shows enterprise buyers increasingly weigh developer velocity and lock-in risk in platform decisions, dynamics VKS 3.8 directly addresses.

Broadcom Expands Kubernetes Platform with VKS 3.8 and CNCF Contribution Read More »

Dstl8 2.0: The AI SRE Built to Replace Observability Platforms

Dstl8 2.0: The AI SRE Built to Replace Observability Platforms

ControlTheory launched Dstl8 2.0 at KubeCon 2026, extending its AI SRE to cover metrics, traces, and GPU-layer visibility without requiring a separate observability platform. The release targets the economics of conventional telemetry ingestion and the cognitive burden of alert-rule maintenance. ECI Research analyst commentary examines what this means for platform engineering teams managing AI infrastructure.

Dstl8 2.0: The AI SRE Built to Replace Observability Platforms Read More »

CloudBolt GA: Bill-Accurate Kubernetes GPU Cost Optimization

CloudBolt GA: Bill-Accurate Kubernetes GPU Cost Optimization

CloudBolt has released GPU optimization in StormForge as generally available, adding bill-accurate GPU cost allocation, node pool reporting, and reliability-aware autonomous rightsizing for Kubernetes environments. The release targets platform engineering and FinOps teams facing simultaneous pressure from finance for GPU cost accountability and from developers for faster, more reliable infrastructure. ECI Research data shows infrastructure provisioning bottlenecks remain a top developer productivity drain, making trustworthy automation a high-value capability.

CloudBolt GA: Bill-Accurate Kubernetes GPU Cost Optimization Read More »

Spacelift Run Observability Brings OpenTofu Trace Analysis to Platform Teams

Spacelift Run Observability Brings OpenTofu Trace Analysis to Platform Teams

Spacelift launched Run Observability at OpenTofu Day North America, making it the first platform to translate OpenTelemetry traces from OpenTofu into structured, actionable findings. The feature identifies which phase, resource, or provider caused a run to fail or slow, eliminating manual log triage. It’s available immediately for Spacelift customers on OpenTofu native stacks.

Spacelift Run Observability Brings OpenTofu Trace Analysis to Platform Teams Read More »

Tigera and Canonical Deliver Enterprise Kubernetes Security Blueprint

Tigera and Canonical Deliver Enterprise Kubernetes Security Blueprint

Tigera and Canonical have partnered to deliver a commercially supported, multi-tenant Kubernetes reference architecture combining Calico’s zero-trust networking with Ubuntu Pro’s long-term OS support. The joint stack targets the policy inconsistency and operational fragility that result when platform teams assemble these layers from separate vendors. For regulated enterprises running hybrid and edge workloads, the combination offers a credible path to consistent security posture and reduced platform maintenance costs.

Tigera and Canonical Deliver Enterprise Kubernetes Security Blueprint Read More »

Portainer-Command: AI Agent Governance for Kubernetes

Portainer-Command: AI Agent Governance for Kubernetes

Portainer has launched Portainer-Command, a governance layer that restricts AI agents to read-only Kubernetes access and routes all proposed changes through Git pull requests requiring human approval. The product targets a growing enterprise need: making AI agent adoption in infrastructure operations auditable, reversible, and controllable. ECI Research data shows that three-quarters of organizations already restrict AI agent use to approved use cases or pilot projects, signaling strong market readiness for enforceable governance tooling.

Portainer-Command: AI Agent Governance for Kubernetes Read More »

Copado Acquires Valo.ai to Close the Runtime Governance Gap

Copado Acquires Valo.ai to Close the Runtime Governance Gap

Copado has acquired Valo.ai to extend its governance platform beyond release management into live production environments. The deal connects operational intelligence, including access visibility and integration mapping, with Copado’s Agentia AI delivery platform. The combined offering targets the growing gap between governed deployments and ungoverned runtime operations in complex Salesforce environments.

Copado Acquires Valo.ai to Close the Runtime Governance Gap Read More »