Office in the Cloud: Metamentor Builds AI Products on Virtual Servers
Russian AI developer Metamentor has moved 80% of its IT infrastructure to cloud provider Timeweb Cloud. The shift has allowed the startup to scale without major investments in buying and installing hardware. The company’s revenue grew from 2 million rubles (about $24,000) to 64 million rubles (about $771,000) in two years.
Metamentor has cut its computing costs by 30%, while reducing server deployment times severalfold. Streamlining routine operations helps the company meet delivery deadlines as its order volume grows.
Virtual Servers Instead of Hardware
Metamentor has connected more than 70 virtual servers in a private network, using cloud databases, S3 storage and automation tools including Terraform and Ansible. The setup allows test and production environments to be launched in minutes. That speeds up the release of AI assistants, voice bots and document-processing systems – the products this young, growing company specializes in building.
“Our core value is sophisticated AI products for businesses. The cloud gives us a reliable foundation on which to build them,” said Danila Dyachenko, CTO of Metamentor.
The company has effectively converted capital expenditures into operating costs. It no longer has to budget for buying hardware, renting space to house it, updating software or covering other needs tied to employees working on-site. Instead, Metamentor pays for the computing capacity it actually uses, scaling resources up or down depending on its workload and number of customers. In practice, that flexibility gives the company a tangible competitive advantage.
The economics of the model show that Russian IaaS and PaaS providers can do more than handle resource-intensive AI workloads – they can also support companies as they grow. That strengthens Russia’s technological independence while accelerating the development of its digital economy.
The Market Needs More Computing Power
Russia’s cloud services market exceeded 416 billion rubles (about $5 billion) in 2025, growing by nearly 30%. Experts expect roughly the same pace of expansion over the next four years, with revenue potentially topping 1 trillion rubles (about $12 billion) by 2030. Demand from AI developers is one of the forces driving that growth.
Further expansion of cloud services and broad adoption of artificial intelligence will also require more data centers. Russia currently has about 200 commercial data centers, but that is not enough to meet demand for computing capacity. “Demand for computing power is growing, particularly as cloud services and AI expand,” said Daniil Novitsky, CEO of the investment group Professionalnye kompleksnye resheniya. Another 42 data centers are scheduled to come online in the next few years, with construction already underway.
Migrating to Cloud Infrastructure
Russian AI companies have been migrating to domestic providers’ cloud infrastructure for several years. In 2024, for example, Robot Miya moved its voice-bot platform to Russian cloud infrastructure geographically distributed across Moscow, St. Petersburg and Kazakhstan. Meanwhile, the Ainergy AI ecosystem moved onto Itglobal.com GPU servers. The international cloud provider, headquartered in St. Petersburg, increased revenue from its GPU Cloud business by 70% in 2025, reflecting industrial demand for powerful graphics processors capable of handling compute-intensive workloads.
Strategic Takeaways and the Path to Scale
Metamentor’s experience shows the effectiveness of a model in which AI developers build on Russian cloud services without taking on major capital expenditures. Over the next two years, this approach is expected to become one of the main drivers behind growth in the number of cloud providers and expansion of the data center market, where demand is being fueled by widespread adoption of generative AI products. The shift from experimenting with generative AI to deploying it at industrial scale is now one of the defining trends in the digital economy.
