“We’re building the hyperscaler for the edge.”
That’s how Dan Wright, CEO of Armada, describes the company’s mission that has now drawn $131 million in new funding from leading venture firms and corporate investors.
The San Francisco-based startup specializes in building mobile, modular AI data centers designed to operate in remote areas with access to unused or underutilized energy. The latest investment round brings in new backers including Pinegrove Capital Partners, Veriten, and Glade Brook Capital Partners, alongside continued support from Founders Fund, Lux Capital, Marlinspike Partners, and Microsoft’s venture fund. Armada declined to disclose its current valuation but confirmed that it has now raised over $200 million since its founding in early 2022.
Leviathan, is Armada’s infrastructure platform, a one-megawatt AI compute unit packaged inside a transportable container. Leviathan offers ten times the compute capacity of Armada’s earlier system and is designed for rapid deployment in areas where traditional data center construction is slow, costly, or impractical.
Source: Armada
“A lot of times with a traditional data center, it takes years. There’s construction, there’s permits, and then you find out you built the wrong thing,” Wright told Bloomberg Television.
Instead, Armada’s systems can be up and running in a matter of weeks. This approach is particularly useful in regions with stranded energy, electricity sources that are available but not fully utilized due to lack of demand or infrastructure. Armada is currently working with Fidelis New Energy and Bakken Energy to deploy Leviathans in energy-producing states such as North Dakota, Texas, West Virginia, and Louisiana.
The company is also extending its footprint internationally. In collaboration with Microsoft, Armada has partnered with Saudi Aramco to install its systems at sites across the Middle East. These installations take advantage of surplus natural gas and other energy sources to support localized AI computing.
“We’ve got about six gigawatts of stranded energy around the country,” Wright said. “You can have a plug-and-play AI factory in weeks. That’s important because China is coming really fast. We need to unlock all that energy.”
Armada’s customer base spans both industry and government. It serves energy companies, mining operations, manufacturers, and logistics providers. On the public sector side, it works closely with the U.S. Department of Defense, including the Navy, which Wright says is exploring the deployment of Armada’s systems across ships and strategic locations.
“It’s kind of nice that the company is called Armada because we’re working closely with the Navy,” Wright said. “We focus on customers that we know will eventually have thousands of what we call connected assets.”
These connected assets include components like Starlink terminals, Skydio drones, SD-WAN routers, and edge compute nodes. Armada brings them all together through its proprietary operating system, the Armada Edge Platform, which provides a single fleet view of assets deployed globally. This allows customers to manage real-time operations from remote and often disconnected environments.
In one example, the Alaska Department of Transportation uses Armada’s platform to reduce latency in drone data processing from over 24 hours to near real-time during avalanche response missions. The shift enables faster decision-making in high-risk, low-connectivity environments.
Wright emphasized that Armada is building a full-stack infrastructure company, similar in concept to hyperscalers like AWS or Azure, but designed for places those platforms do not reach. While cloud giants focus on large-scale data centers in major metro regions, Armada is targeting the other 70 percent of the world that lacks such access.
The company’s infrastructure is built with mobility and flexibility at its core. Leviathan units can be delivered directly to energy sites, eliminating the need for new pipelines or expensive transmission networks. This has drawn attention from customers seeking to reduce their emissions from flare gas or monetize surplus power through compute workloads.
“Some energy companies have flare gas they’re currently releasing into the atmosphere,” Wright said. “They’re under pressure to do something with it. Now they’re using turbines to convert it into electricity and power infrastructure.”
Connectivity is a critical enabler of this vision. Armada has partnered with Starlink since its earliest days, and Wright said many customers now rely on satellite internet not just for backup, but as their primary connection. Starlink’s continued satellite launches and improved service tiers have made it viable for enterprise-grade applications.
“Most customers don’t want to send all their data back. They only want to send metadata,” Wright explained. “That’s where the synergy is between us and Starlink. We handle local data processing, and Starlink sends back the metadata.”
This localized computing model supports real-time inference, federated learning, and low-latency decision-making. Armada started by focusing on inference at the edge and is now evolving into a platform that supports full-scale AI training and model refinement.
A significant part of Armada’s go-to-market strategy is tied to its relationship with Microsoft. Customers already using Microsoft Azure in connected environments can extend those same capabilities to remote locations using Armada’s infrastructure. This includes services like Azure Stack, as well as production automation tools developed in partnership with Microsoft and Halliburton for the energy sector.
“They don’t have to rebuild everything to run on Linux or CUDA,” Wright said. “They can assume they’ll have access to the same Azure stack at the edge. That’s the power we’re delivering.”
With its growing list of commercial and government partners, Armada aims to become the edge infrastructure provider of record for organizations operating at global scale.
“We’re working with some of the largest companies in the world,” Wright said. “What AWS is for metro areas, we want to be for the rest of the world.”
