A Naples-based startup that builds infrastructure software for managing Kubernetes clusters at scale has closed a seed round of €2.9 million. Clastix announced Tuesday that CDP Venture Capital, Italy's national innovation fund, led the financing, joined by French AI lab Mistral AI and Vertis SGR.
The timing is deliberate. Clastix shipped the first production version of kMetal, its commercial bare-metal Kubernetes platform, just weeks ago—a release that packages years of open-source work into a product designed to manage sprawling AI training and inference clusters without the overhead of traditional virtualization layers. That bet on what the company calls "AI-scale infrastructure" persuaded Mistral to invest as both customer and strategic backer, a vote of confidence that carries weight in a market crowded with Kubernetes tooling.
"Clastix is the backbone of our Kubernetes infrastructure," said Timothée Lacroix, Mistral's co-founder and CTO, in a statement. "Their work on multi-tenancy and managed control planes helps us to enable our training and inference clusters at scale, with the efficiency that frontier AI demands. By backing Clastix, we're investing in Europe's AI stack."
The company's approach centers on a technical shift: running Kubernetes control planes as pods inside a shared management cluster rather than allocating dedicated infrastructure for each tenant. It's a pattern that OVHcloud, the French cloud provider, has adopted for its managed Kubernetes service, according to a Cloud Native Computing Foundation case study. NVIDIA, too, has deployed Clastix's open-source Kamaji project to orchestrate hundreds of control planes for its DOCA platform running on ARM-based BlueField data processing units, according to a Clastix case study published in April 2025.
Clastix maintains two open-source projects under Apache and CNCF governance. Kamaji handles control plane management through kubeadm. Capsule, which the company donated to the CNCF Sandbox in late 2022, provides workload isolation inside shared clusters. The commercial kMetal platform layers KVM-based worker virtual machines, GPU pooling, and per-tenant VPC networking on top of both, according to product documentation—essentially offering fleet operations without requiring a separate hypervisor.
The customer roster extends beyond cloud infrastructure. Italian telecom operator Fastweb has used Capsule to deliver multi-tenant Kubernetes-as-a-Service to hundreds of customers on its FASTcloud platform, a deployment detailed in a case study published February 17, 2022.
Vertis SGR participated in the round through Vertis Venture 6 Digital Sud, a €60 million fund targeting Southern Italy startups. The vehicle draws financing in part from the EU's NextGenerationEU recovery program and is anchored by CDP's Digital Transition fund. That structure reflects Italy's push to channel European recovery capital into regional tech ecosystems, though the fund launched relatively recently and has yet to build an extensive portfolio track record.
Alessandro Scortecci, chief investment officer for direct investments at CDP Venture Capital, positioned the deal around infrastructure demands tied to what some are calling the physical AI revolution—systems that require resilience and scale beyond traditional cloud workloads.
CEO Adriano Pezzuto struck a similar note, though with broader ambition. "The next phase of AI won't be decided by models alone but by who controls the infrastructure layer beneath them," he said. "This funding goes toward building that layer for what's coming."
Clastix plans to expand hiring across customer success, solutions engineering, and go-to-market functions. Recent LinkedIn postings from Pezzuto list openings for a sales director, head of marketing, presales engineer, and several engineering roles. The company will continue investing in Kamaji's open-source development alongside accelerating kMetal's roadmap and funding research into AI-focused workload patterns.
The company did not disclose valuation. For Clastix, which took no outside capital before this round, the seed marks a shift from pure open-source maintenance to supporting a commercial product—one shaped by the same forces that have drawn enterprise and venture attention to infrastructure layers generally: the compute intensity of modern AI, the economics of multi-tenancy, and the persistent complexity of running Kubernetes at any meaningful scale.
