Factory-Built AI Data Center Startup Mavria Raises $100 Million Seed At $1 Billion Valuation

Mavria on November 9, 2026 unveiled its factory-built AI data-center business and disclosed $130 million raised to date across pre-seed and seed rounds. The total includes a new $100 million round led by Paradigm at a $1 billion valuation.

AMP PBC, Liquid 2, Critical Frontier and Dream Ventures also participated in the latest round. Mavria said it is already putting the capital to work securing critical supply chains, scaling manufacturing and expanding its team.

The announcement also introduces Terrain, the company’s first product. Chief executive Adam Blake said Mavria was formed to address a mismatch between the AI industry’s compute ambitions and the industrial capacity needed to build data centers and their supporting power infrastructure. The company’s response is to replace custom field construction with repeatable manufacturing, treating infrastructure as a product whose design and production can improve from one project to the next.

Terrain Integrates Power, Cooling and IT Systems

Mavria describes Terrain as a factory-built data-center platform for high-density AI compute, organized in standardized 5 MW blocks. Power, cooling, controls and IT pods are designed, integrated and tested as a standardized system before deployment.

That integration is central to Mavria’s commercial pitch. Instead of a customer sourcing separate modules and coordinating their assembly, Terrain is intended to provide an integrated, factory-tested product. Mavria says the approach is designed to bring compute online faster, with more predictable delivery and lower costs, although its initial product materials do not provide cost or schedule measurements.

The intended customers are frontier labs, neoclouds, developers and infrastructure owners. Their compute would run inside Terrain, while Mavria supplies the physical infrastructure needed to operate it. The company explicitly says it does not operate a cloud or sell compute, making its business distinct from providers that rent AI processing capacity.

Terrain 01 is being built in Houston as the first physical demonstration of the approach. Mavria describes it as in production, but full product details remain forthcoming. No separate commercial availability date was announced.

Manufacturing Capacity Is Central to the Model

Mavria’s operating model standardizes how infrastructure is developed, built and deployed through repeatable engineering, procurement and manufacturing. Development, procurement and manufacturing advance in parallel, while critical equipment, supply chains and production capacity are secured ahead of demand. The company says this is intended to reduce bottlenecks when projects are ready.

Writing after the announcement, co-founder Seth Cohen said Mavria was founded in June 2026 after the founders’ work inside government began in 2025. He set a target of deploying Terrain in weeks rather than months or years and linked faster deployment to less disruption for host communities. His longer-term ambition is for Mavria to manufacture products upstream from Terrain once it proves itself in data centers, extending the company beyond its first product.

The three founders, Blake, Nick Poradek and Cohen, previously worked together at the U.S. Department of Energy. Company biographies say Blake built more than 1,000 energy projects, sold a software business to Vista Equity Partners and advised the secretary of energy. Poradek sold a manufacturing company to Liberty Energy (LBRT ) and then helped build Liberty’s manufacturing platform, while Cohen’s background spans nuclear policy, government and capital. The broader team brings experience in data-center engineering at NVIDIA and AWS, large-scale modular LNG manufacturing and infrastructure project finance.

For Mavria, the central question is execution: whether Terrain 01 can demonstrate that standardized, factory-tested data centers can be produced repeatedly and meet the company’s goals for deployment speed, delivery predictability and lower costs.