Clastix's Open Source Wedge Lands NVIDIA and Mistral as Customers

The Naples-based startup is betting its open-source projects Capsule and Kamaji are the path to selling enterprises a simpler way to run Kubernetes at scale.

About Clastix

Published

The first thing you notice is the quiet. You are looking at a dashboard that shows not one, not ten, but hundreds of Kubernetes control planes. In a conventional setup, this would be a cacophony of independent virtual machines, each with its own operating system, its own lifecycle, its own license fee. Here, they are just pods. Neat rows of them, running on shared metal, managed as a single fleet. This is the view from inside kMetal, Clastix’s commercial platform, and it answers a question that only emerges when you are already in deep: what does infrastructure look like after you have won the argument for Kubernetes, and now have to live with the consequences of your own success?

Founded in Naples in 2020 by Adriano Pezzuto and Dario Tranchitella, Clastix didn’t start with a commercial product. It started with an itch, the kind engineers get after managing sprawling Kubernetes deployments for large enterprises. Their first moves were open-source projects: Capsule, for multi-tenancy, and Kamaji, for running tenant control planes as hosted services. These were tools built to solve the operational complexity that comes with scale, and they found an audience. According to the company, infrastructure teams at NVIDIA, Mistral, OVH, and Rackspace adopted Kamaji, using it to run hundreds to thousands of tenant clusters in production [Introducing kMetal: From Hosted Control Planes to Bare Metal | CLASTIX]. That adoption became the wedge for kMetal, the paid platform that wraps those open-source cores with enterprise-grade orchestration, support, and a promise of radical simplification.

The Wedge of Shared Control Planes

Clastix’s bet is architectural, not just incremental. Traditional Kubernetes-at-scale often involves a stack of licensed virtualization software (like VMware), the etcd datastore for each cluster, and the independent lifecycle management of each control plane. kMetal proposes to collapse those three stacks into one Kubernetes-native platform [iH CLASTIX CLASTIX kMetal Technical Whitepaper Three Stacks Into One:]. The mechanism is Kamaji. It allows the control plane for each tenant cluster,the brain of the operation,to run not as a separate virtual machine, but as a pod on a shared, underlying Kubernetes cluster. This is the technical heart of the quiet dashboard: control-plane consolidation. It aims to free up existing machines to be redeployed as worker nodes, cut hypervisor licensing costs, and turn fleet operations from a manual slog into a software-defined workflow [Introducing kMetal 1.0 - for Helping Enterprises Control & Scale Kubernetes More Efficiently | CLASTIX].

The target is the infrastructure team inside large enterprises in AI, telecommunications, cloud services, and the public sector, particularly in Europe. These are organizations where Kubernetes has moved from experiment to essential, and where the bill for complexity is now coming due. kMetal is not another application-layer tool; it is a foundational layer, competing with the need to hire more platform engineers or write more custom automation.

Fuel for the AI Infrastructure Layer

The timing of Clastix’s recent seed round reads as a validation of this thesis. In September 2026, the company closed a €2.9 million (approximately $3.1 million) seed round led by CDP Venture Capital Sgr, with participation from AI lab Mistral and venture firm VERTIS SGR [Vestbee, September 2026]. The investor mix is telling. CDP Venture Capital is a major Italian venture arm, signaling domestic support. Mistral’s participation is more than financial; as a named customer, its involvement suggests a strategic alignment with Clastix’s technology as part of the compute stack needed for large-scale AI workloads [Clastix, October 2026]. The company explicitly frames the capital as fuel to "build AI’s infrastructure layer" [Clastix, October 2026].

The funding is earmarked for product development, customer success, solutions engineering, and go-to-market expansion [Vestbee, September 2026]. The current team is small, reported at 1-10 employees [Clastix - 2025 Company Profile, Team & Competitors - Tracxn, 2024], but hiring is underway. Open roles point to the next phase: a Sales Engineer, an Account Executive, and a Technical Content role, all remote across Europe [Perplexity Sonar Pro Brief]. This is the build-out from an engineering-led open-source project to a commercial entity with a sales motion.

Founder Role Background
Adriano Pezzuto Co-founder, CEO Nearly three decades in IT; based in Milan [Adriano Pezzuto - CLASTIX
Dario Tranchitella Co-founder, CTO Engineer with deep Kubernetes and infrastructure experience [Perplexity Sonar Pro Brief].
Dimitri Bellini Co-founder Co-founded Clastix with Tranchitella [Dimitri Bellini - Quadrata

The Inevitable Friction of Platform Shifts

For all its elegant architecture, Clastix’s path is not without friction. The company is asking enterprises to make a foundational platform change, a category of purchase that involves long evaluation cycles, deep technical due diligence, and organizational buy-in. While the open-source projects provide a low-friction entry point, converting that usage into a commercial kMetal deal is a separate challenge. Furthermore, the competitive landscape, while not named in sources, is implicit. They are competing against in-house platform teams building similar tooling, against the managed Kubernetes services of large cloud providers, and against other commercial Kubernetes distribution vendors. Clastix’s answer lies in its specificity: it is not selling a general-purpose distribution, but a system optimized for one very painful, very expensive problem,fleet-scale multi-tenancy.

The early customer logos provide a counter-argument to the friction. Landing NVIDIA and Mistral as production users of its open-source core is a powerful signal of technical credibility. These are organizations with extreme performance and scale demands. Their adoption suggests Kamaji solves a real problem for sophisticated operators. The commercial bet is that these same organizations, and others like them, will pay for the consolidated management, support, and additional enterprise features of kMetal to manage that scale with fewer people and lower cost.

What to Watch in Naples

The next twelve months will be about proving the wedge works. The key milestones are less about new features and more about commercial traction and organizational scale.

  • First Major kMetal Deals. The seed round was closed in September 2026. The capital will be used to build the sales function. The true test will be the announcement of first enterprise contracts for the kMetal platform, moving beyond open-source adoption.
  • European Expansion. The company is based in Naples and has strong Italian backing. Its open-source users, however, span Europe and North America. Watch for hiring and customer announcements that signal a deliberate expansion beyond its home region.
  • The AI Workload Proof Point. With Mistral as both investor and customer, Clastix is positioned as infrastructure for the AI boom. Concrete case studies showing kMetal managing large fleets of GPU-accelerated Kubernetes clusters would powerfully validate this positioning.

The quiet of the dashboard, then, is the point. It represents a vision of infrastructure where complexity is abstracted, not accumulated. Clastix is betting that the next wave of enterprise Kubernetes isn’t about getting it running, but about making the sheer, overwhelming fact of it disappear into the background, as reliable and unnoticed as electricity. The cultural question kMetal implicitly answers is one of fatigue: what do we build when we are tired of managing our own tools, and just want them to work?

Sources

  1. [Vestbee, September 2026] Clastix secures €2.9M to scale its Kubernetes infrastructure platform | https://www.vestbee.com/insights/articles/clastix-secures-2-9-m
  2. [Clastix, October 2026] Why We Received €2.9M to Build AI’s Infrastructure Layer | https://clastix.io/post/why-we-received-eur-2-9-million-seed-funding-to-build-ai-infrastructure-layer/
  3. [Introducing kMetal: From Hosted Control Planes to Bare Metal | CLASTIX] Introducing kMetal: From Hosted Control Planes to Bare Metal | https://clastix.io/post/introducing-k-metal-from-hosted-control-planes-to-bare-metal/
  4. [iH CLASTIX CLASTIX kMetal Technical Whitepaper Three Stacks Into One:] iH CLASTIX CLASTIX kMetal Technical Whitepaper Three Stacks Into One: | https://media.trustradius.com/product-downloadables/ET/OX/T5EBUPEN7OT5.pdf
  5. [Introducing kMetal 1.0 - for Helping Enterprises Control & Scale Kubernetes More Efficiently | CLASTIX] Introducing kMetal 1.0 - for Helping Enterprises Control & Scale Kubernetes More Efficiently | https://clastix.io/post/introducing-k-metal-1-0-for-helping-enterprises-control-and-scale-kubernetes-more-efficiently/
  6. [Clastix - 2025 Company Profile, Team & Competitors - Tracxn, 2024] Clastix - 2025 Company Profile, Team & Competitors - Tracxn | https://tracxn.com/d/companies/clastix/__k7u5wGIVDqOmwmy6V4xQZ_nStlttO3IV0qBZ5512plo
  7. [Adriano Pezzuto - CLASTIX | LinkedIn, 2026] Adriano Pezzuto - CLASTIX | LinkedIn | https://www.linkedin.com/in/adrianopezzuto/
  8. [Dimitri Bellini - Quadrata | LinkedIn, 2026] Dimitri Bellini - Quadrata | LinkedIn | https://www.linkedin.com/in/dimitri-bellini/
  9. [Perplexity Sonar Pro Brief] Perplexity Sonar Pro Brief
  10. [Finsmes, September 2026] Clastix Raises €2.9M in Seed Funding | https://www.finsmes.com/2026/09/clastix-raises-e2-9m-in-seed-funding.html

Read on Startuply.vc