Nebul Deploys Mirantis' k0rdent to Power Scalable AI Inference Services

Mirantis, a leader in open-source cloud platforms, has announced that Dutch private cloud provider Nebul has adopted its k0rdent solution to deliver scalable, on-demand AI inference services. This deployment allows Nebul to execute high-performance, low-latency AI applications on their NVIDIA-powered infrastructure.
Nebul, an NVIDIA Elite and Cloud Partner, is known for its activities in high-performance computing, artificial intelligence, and machine learning. With k0rdent, the firm is now providing AI Inference-as-a-Service, giving customers the power to deploy trained models to new data with increased speed, security, and efficiency.
Open-Source Infrastructure to Support AI at Scale
k0rdent is a Kubernetes-native multi-cluster management platform that integrates tightly with NVIDIA GPU Operator and Gcore’s Everywhere Inference solution. This setup allows Nebul to manage distributed AI inference workloads efficiently, using automated provisioning and centralized policy controls to optimize GPU use across their infrastructure.
Alex Freedland, co-founder and CEO, Mirantis, the maintainer of k0rdent, said, “We believe open source is the enabler for infrastructure to support AI. Nebul is demonstrating the enormous potential of open technologies to solve one of the most complex challenges in IT today — delivering AI workloads reliably at scale.”
Also read: Hyperscale Data Secures Up to $50M in Preferred Stock Financing to Accelerate Michigan Data Center Buildout
Streamlining AI Infrastructure for Enterprise Use
Seva Vayner, the product director of Edge Cloud and Edge AI at Gcore, explained that Nebul is showcasing the need for infrastructure capable of dynamically adapting to the needs of end customers, ensuring that AI applications meet business requirements in terms of both cost and performance.
He mentioned that their Everywhere Inference product offers an exceptionally simple portal to deploy and manage AI inference, while the Smart Routing feature ensures that inference tasks are directed to the nearest GPUs to minimize latency.
Launched in early 2025, k0rdent is designed to help platform engineers reduce infrastructure sprawl and operational complexity across diverse environments, including cloud, on-prem, and edge deployments. It leverages open-source Cluster API to manage Kubernetes clusters with declarative automation and production-ready templates.