Copenhagen-based startup Velatir has locked down €5 million in seed funding to push its AI governance platform deeper into European markets and scale its workforce.
The capital arrives just six months after a pre-seed round and follows the company’s first commercial rollout. Investors see urgency here. Businesses keep piling on AI tools, yet few have any real visibility into how employees and autonomous agents actually use them.
Spintop Ventures and Ugly Duckling Ventures co-led the round. Norrsken Evolve joined in, along with angels Jan Oberhauser from n8n and Thomas Visti from Universal Robots and MiR. The Danish Export and Investment Fund added a matching loan.
Velatir, founded by Michael Sørensen, Elias Sørensen, Christian Møller and Andreas Paulli, builds infrastructure that helps organizations track AI activity across browsers, devices, vendors and internal systems. The platform unifies security enforcement, usage monitoring and policy controls in one place. Companies get real-time data on costs, data flows and how mature each AI use case has become. A built-in catalogue covers more than 4,000 AI tools.
The founding team sees a governance gap widening as AI embeds itself into daily workflows. Compliance questions multiply. Where does data go? Who uses which tools? Does any of it violate internal rules or regulatory requirements? Velatir answers those questions through centralized guardrails.
Perhaps most notable: the entire platform runs on European-owned and hosted infrastructure. No US hyperscalers.
“Sovereign cloud is a common phrase, and it often means American platforms with a European label,” said COO Christian Møller. “We built Velatir on European-owned and hosted infrastructure from the very beginning. It is the harder path, but the only viable one for us.”
That positioning matters as European regulators and enterprises grow more cautious about where AI infrastructure and data physically reside. Demand appears to be rising accordingly. The fresh capital will fund expansion and hiring across the continent.
