The race to build AI infrastructure has never been more competitive — or more critical.

With a recent multi-billion-dollar deal with Microsoft Corp. and deep roots in distributed systems, Nebius B.V. is emerging as a central player in the evolution of next-generation data centers — essentially, AI factories. How is Nebius advancing its goal to combine the efficiency of supercomputing with the ease of cloud computing?

“The core business we built is actually the cloud,” said Roman Chernin (pictured), co-founder and chief business officer of Nebius. “We build a multi-tenant cloud with a very sophisticated software platform with a lot of managed services, and we build it to serve a variety of customers, all the way from small start-ups to the large enterprises. We can speak about it separately, like how we observe the evolution of the use cases, how we observe the evolution of the customers, and who is driving the consumption now, and so on.”

Chernin spoke with theCUBE’s John Furrier at theCUBE + NYSE Wired: AI Factories – Data Centers of the Future event, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed how Nebius is positioning itself as a builder of the AI infrastructure that will power the next generation of applications.

Nebius’ AI infrastructure prizes specialization as its competitive edge

Unlike general-purpose cloud providers, Nebius is purpose-built for AI workloads. That specialization is the company’s secret sauce, according to Chernin. By optimizing infrastructure and developer tools specifically for large-scale training, inference and data-intensive applications, Nebius cuts deployment times dramatically — from months to hours in many cases.

“We don’t try to build any universal cloud,” Chernin said. “We go only after AI-centric workloads, meaning large distributed training, inference on scale, and we build the platform from the ground up to actually address these use cases in the best performance. Then the approach that we took is that it’s not just managed infrastructure, but the cloud actually addresses a lot of developer pain points.”

This specialization resonates with a wide spectrum of clients. Foundational model builders led the first wave of adoption, followed by vertical AI startups — and subsequently, enterprises such as Shopify are beginning to engage, according to Chernin. Over time, traditional industries from healthcare to finance are expected to drive growth, demanding secure, hybrid-ready platforms that blend cloud flexibility with enterprise control.

“We already work with Shopify and other tech-savvy, tech-advanced companies like them,” Chernin said. “We’ll see the classical enterprises coming to the table as well. Those types of customers don’t look just for bare metal infrastructure … they need platforms, they need a developer platform, they need data platforms with all the security and all the classical cloud stuff.”

For Nebius, the bigger vision is building the next layer of AI platforms. Several opportunities are arising in inference, fine-tuning, reinforcement learning and developer tooling — all designed to help enterprises apply proprietary data to open-source models, according to Chernin. For organizations wary of locking into a single provider such as Google LLC or Anthropic PBC, Nebius aims to lower the barrier to building intelligent systems on open frameworks.

“We want to help people build on open-source models first of all,” Chernin said. “At the end of the day, if people decide to build on Gemini, they will build in the Google ecosystem. But what if you want to build on open source? What if you need to reduce the latency, reduce the price and apply the data more extensively and extract more value?”

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of theCUBE + NYSE Wired: AI Factories – Data Centers of the Future event:

Photo: SiliconANGLE