Kubernetes has moved well past its early adoption phase. The new Komodor 2025 Enterprise Kubernetes Report shows that technical teams are shifting their focus from running containers to managing a growing mix of AI workloads and advanced automation practices like GitOps.
“Organizations have made Kubernetes their standard, but our report shows the real challenge is operational, not architectural,” said Itiel Shwartz, CTO of Komodor. “Even as practices like GitOps and platform engineering gain traction, enterprises still grapple with change management, cost control, and skills gaps. At the same time, the growth of AI/ML workloads and AIOps marks the next frontier, reinforcing Kubernetes as the backbone of enterprise infrastructure.”
AI/ML becomes a core workload
More than half of the organizations in the study are already running some form of AI or ML workloads on Kubernetes. These range from model training and experimentation to real-time inference.
The move makes sense. Kubernetes provides flexible scaling and resource management, both of which are essential for GPU-heavy workloads. However, running AI on Kubernetes introduces new challenges. Scheduling GPU resources efficiently has become a concern, with many teams reporting underutilized hardware. Solutions such as queue-based scheduling and Kubernetes-native GPU operators are emerging to address this problem.
More than 40% of organizations plan to expand their use of orchestration and scheduling tools to better manage GPUs. This trend shows that Kubernetes is becoming the default platform for production AI workloads, not just traditional applications.
GitOps takes center stage
As environments grow, so does the need for reliable, repeatable deployment practices. The report highlights a rise in GitOps adoption, with a large majority of teams now using it to manage Kubernetes configurations. GitOps brings version control and automation to infrastructure changes, making rollbacks and drift detection easier.
ArgoCD is the most widely used GitOps tool, with Flux also seeing strong adoption. Together, they form the backbone of many Kubernetes deployment pipelines. By standardizing on GitOps, teams are reducing errors and improving stability, which is especially important in multi-cluster and hybrid environments.
Namespaces remain the most common method for isolating applications, but many teams are also turning to separate clusters for added security and control. This combination helps organizations balance risk and operational overhead.
Complexity drives platform engineering
With Kubernetes clusters multiplying across on-premises data centers, public clouds, and even edge locations, many companies are forming dedicated platform engineering teams. These groups focus on building internal developer platforms and managing the tools that keep Kubernetes environments consistent and secure.
This shift is driven by the need to reduce fragmentation. The report notes that organizations with unified policies and standardized tools experience fewer outages and lower operational costs. Without this level of coordination, teams often face tool sprawl and inconsistent practices.
