In Conversation with Dan Wright, Cofounder & CEO of Armada ($2B neo-hyperscaler for the edge)

Episode Summary

Executive Summary: Dan Wright explains Armada’s thesis: the next wave of AI will happen at the remote edge, not just in cloud back offices. Drawing on experience at AppDynamics and DataRobot, he saw that connectivity, security, and data-sovereignty barriers kept critical industries from using AI on-site. Armada builds full-stack edge infrastructure to bring compute, models, and software to data in energy, mining, manufacturing, defense, and emergency response.

Main Topics: Founding thesis: AI at the remote edge (Priority: 5/5): Wright says Armada was founded on the belief that AI must be deployed where data is generated, especially in remote industrial settings where cloud use is constrained by latency, sovereignty, and security. Why legacy approaches failed (Priority: 5/5): Before Armada, companies used siloed DIY setups like GPUs in closets or rig houses, which were unsafe, hard to scale, and limited to single-use cases. Full-stack platform and marketplace model (Priority: 5/5): Armada combines modular hardware, an operating system (AEP), and a software marketplace to let customers run models locally, fine-tune them securely, and even commercialize proprietary models. Customer segments and use cases (Priority: 4/5): The company serves large commercial customers and public sector organizations, including energy, mining, manufacturing, the U.S. Navy, emergency response, and sovereign AI initiatives. Growth drivers and market tailwinds (Priority: 4/5): Wright attributes hypergrowth to rising global connectivity, exploding edge data from sensors/drones/robotics, and broader adoption of AI tools that enterprises now want in operations. Culture, hiring, and execution discipline (Priority: 4/5): Armada emphasizes a strict mission-driven culture, highly selective hiring, and ongoing performance evaluation to maintain speed and quality during rapid scaling. Global expansion and future outlook (Priority: 4/5): Armada is expanding across the Americas, Middle East, Europe, and Asia-Pacific, with deployments that can scale from megawatts to gigawatts and support sovereign AI goals.

Key Arguments: AI’s highest-value use cases in industrial sectors require processing data locally at the edge, not sending it to distant clouds. Connectivity alone is not enough; without local infrastructure and secure deployment, edge AI cannot be operationalized. A unified full-stack platform enables repeated deployment, model improvement, federated learning, and a broader ecosystem of third-party builders. Industrial customers need speed, scale, and sovereignty: rapid deployment, elastic capacity, and control over sensitive data. Data becomes dramatically less valuable as time passes, so real-time edge processing unlocks operational, safety, and cost benefits. Armada’s platform can turn proprietary customer models into monetizable assets through private and public marketplaces. The company believes the edge market is larger than the hyperscale cloud market because roughly 70% of the world lacks hyperscale data centers. Strong culture and unusually selective hiring are central to executing a technically complex, high-stakes mission.

Data Points: Company age at time of interview: a little over 3 years - Wright says Armada started in end of 2022 and is now a little over three years old. Previous company acquisition valuation: $44 billion - AppDynamics was bought by Cisco shortly before IPO. Current team size: about 400 employees - Wright says Armada has around 400 employees globally. Recent capital raised: about $230 million - He references a recent round of roughly $230M. Global data-center coverage: about 30% of the world - He says only about 30% of the world has big hyperscale data centers. Remote edge coverage: about 70% of the world - Armada targets the part of the world without hyperscale data centers. Edge data generation by 2028: 80% of data - He cites Gartner saying most data will be generated at the edge by 2028. Alaska latency reduction: 28 hours to real time - For emergency response drone data processing in Alaska. Navy deployment: first modular data center on a Navy ship last summer - Wright says Armada was first to deploy this on a Navy ship. Australia deployment: initially 10 megawatts - WinDC deployment starts at 10 MW with plans to scale much larger.

Pivotal Quotes: "we can bridge the digital divide. That's the mission of the company and build the hyperscaler for the edge." — Dan Wright: Defines Armada’s core mission and market position. "the three S's: speed, scale, and sovereignty." — Dan Wright: Summarizes the customer value proposition. "We make the computers do the work" — Dan Wright’s customer example: Describes automating manual inspection and reporting tasks in energy operations.

Implications: Armada is betting that industrial AI will move from centralized clouds to sovereign, edge-based infrastructure. If that thesis holds, the winners will be platforms that combine hardware, software, and ecosystems to serve remote operations globally.