Driving 40% lower Kubernetes operating costs with a multi-environment platform

Sworn to secrecy

This organization is a large healthcare provider supporting critical clinical and operational workloads across hospitals, clinics, and enterprise IT environments.

Shared anonymously at the customer’s request, this story reflects a real-world journey to bring structure and consistency to Kubernetes operations, while maintaining the reliability and governance required in healthcare environments.

While identifying details have been removed, the challenges faced, decisions made, and results achieved are real.

At a glance

• Industry: Healthcare
• Footprint: Large, multi-region healthcare organization operating hospitals, clinics, and enterprise IT platforms
• Objective: Establish a Kubernetes operating model that could scale across multiple environments without disrupting clinical systems  
• Key use cases: Centralized Kubernetes provisioning, upgrades, and lifecycle management across on-premises and cloud environments  
• Solution: Spectro Cloud Palette, SaaS
• Outcome: A standardized Kubernetes platform delivering reliable 99.9% uptime and driving toward up to 40% lower operating costs in the first year

A Kubernetes journey from pilot to priority

For years, this large healthcare organization relied on a virtualization-first model. Core systems ran on VMware, supported by teams with deep operational expertise, while containerized applications were deployed using Docker Swarm — a setup that had been in place for nearly a decade.

By 2025, the organization began modernizing its platform, experimenting with Kubernetes across Azure Kubernetes Service (AKS) and on-premises environments, where Mirantis Kubernetes Engine (MKE) was used to run Kubernetes on vSphere. As teams moved beyond pilots and looked to accelerate adoption, the limits of its existing operating model quickly became clear.

Managing clusters individually across cloud and on-premises environments meant using largely manual processes. There was no unified management plane and no consistent approach to provisioning, upgrades, or lifecycle management, with each environment requiring its own operational playbook. 

Terraform, while powerful, had also become difficult to manage at scale, with dozens of resource blocks per cluster slowing the team down. At the same time, Kubernetes expertise remained limited, placing growing operational pressure on a small number of engineers.

VMware continued to play a strategic role, supporting critical systems operated by large, highly skilled teams. While leadership was concerned about Broadcom’s pricing changes, they were also nervous of the risk involved in retraining teams or forcing a rapid shift away from VMware. They needed flexibility over time, without disrupting existing operations.

Despite these fears, the direction was clear and the organization expected Kubernetes to grow from a handful of clusters into a standardized platform used across hospitals and clinics. To get there, they needed a more consistent, scalable operating model that could both provide flexibility across environments and respect its VMware reality.

Solving both operational and business needs

The platform team was already supporting a mixed estate, and they knew any solution had to work within that reality, not replace it. Knowing Mirantis couldn't scale with them, the team evaluated other Kubernetes management platforms, including Red Hat OpenShift, but it also didn’t meet the goal of unifying existing and future deployments under a single operating model.

With Spectro Cloud Palette, the team found a platform that fit both current and future requirements.

From day one, Palette could manage their existing Kubernetes clusters running on vSphere and AKS from a single control plane, while preserving flexibility to expand to Amazon EKS over time. Automation from cluster provisioning through Day-2 operations made it possible to scale Kubernetes adoption without overwhelming the platform team.

Palette also addressed leadership concerns around security, compliance, and long-term risk. It supported healthcare-grade requirements, including HIPAA-aligned operations, BAA obligations, and centralized, auditable policy enforcement across environments. And its worker-node-only pricing model and ability to support multiple environments under a single platform gave leadership confidence in the economics of their Kubernetes strategy, allowing them to control costs, avoid lock-in, and make gradual infrastructure decisions over time.

Higher flexibility, lower costs

By the end of 2025, leadership could already see early results from unifying Kubernetes management through Palette, across how clusters were deployed, operated, and governed. Standardized workflows replaced previously fragmented processes, reducing manual effort and easing the operational burden on the platform team.

These simplified operations also point to meaningful efficiency gains, with the organization tracking toward its goal of reducing Kubernetes operating costs by 40% in the first year.

Cluster availability has improved as automated lifecycle management and upgrades reduce the risk of configuration drift and unplanned downtime. Early production environments are meeting uptime targets of 99.9% across managed clusters, supporting the reliability expectations of clinician- and patient-facing services.

Consistent provisioning and GitOps-based workflows are improving predictability across environments, allowing teams to bring new clusters online with greater confidence while maintaining governance and compliance controls.

As the platform team introduced Amazon EKS alongside existing Azure and on-premises environments, the organization is already benefiting from increased flexibility in workload placement. They manage EKS using the same lifecycle, governance, and operational model applied elsewhere, eliminating cloud-specific processes and improving visibility across the Kubernetes estate.

Taken together, these early results are reinforcing confidence in the organization’s multi-cloud strategy and establishing a strong foundation for continued scale and expansion.

What's next on the journey

Rather than treating Kubernetes as a series of isolated projects, the organization focused early on establishing a consistent operating model for some of its most critical systems. That foundation made it easier to expand Kubernetes adoption without introducing new operational silos.

Building on that progress, the platform team plans to deepen its investment in AWS while extending Kubernetes adoption across additional use cases. With Palette as a long-term platform foundation, they can support a broader range of data-intensive workloads — including emerging analytics and AI-driven use cases — while maintaining consistent governance, security, and operational control as their cloud strategy evolves.

At the same time, the organization is reassessing its virtualization strategy in response to ongoing cost and operational pressures. As part of that effort, the platform team is evaluating how Palette’s Virtual Machine Orchestrator could support running virtual machine workloads alongside Kubernetes on shared infrastructure, creating flexibility to reduce long-term reliance on parallel platforms over time.

With a standardized platform in place, the team is well positioned to support today’s clinical and operational needs while retaining the control and flexibility required to adapt as requirements change.