Defense Unicorns co-founder on modernizing software delivery in air-gapped, mission-critical military environments

Summary

Defense Unicorns just closed a $136M Series B led by Bain Capital, hitting unicorn status — a milestone that carries extra weight given the company literally named itself after the outcome.

Rob Slaughter, co-founder and CEO, spent 13 years in the military before starting the company five years ago with fellow veterans. The pitch is simple: the U.S. military has a severe software delivery problem, and the people best positioned to fix it are the ones who lived it.

The problem

Three constraints define military software environments. First, most systems are air-gapped — fully disconnected, semi-disconnected, or locked behind firewalls so tight that tools like GitHub are inaccessible. Second, cybersecurity accreditation standards are exceptionally demanding. Third, the people actually operating these systems are active-duty soldiers, sailors, airmen, and marines whose primary job is combat, not IT.

The hardware situation compounds all of this. In military systems, a software update often requires a hardware update too. When new programs fail to deliver — which happens repeatedly — the hardware gets frozen in place. Slaughter describes programs that stalled two or three times, leaving warfighters running systems built on Windows 95-era infrastructure.

The platform

Defense Unicorns built an open-source baseline called UDS that lets software be packaged and deployed across unclassified environments, classified environments, and edge weapon systems — F-16s, F-22s, and submarines. Their largest customer is the Department of the Navy, specifically the submarine community.

The core capability is on-demand software updates on constrained hardware. Slaughter's example is concrete: a pilot might need one specialized local AI model for a mission on Monday and a completely different one on Tuesday. Under current practice, that swap would require taking the aircraft offline for a full maintenance period. Defense Unicorns makes that a software operation instead.

The AI constraint

Running modern AI on legacy military hardware is a real ceiling. Slaughter's answer is not to wait for hardware refreshes — it's to find systems that received a hardware update one or two years ago, which may not be cutting-edge but can run significantly more software than what existed when they were fielded. The ability to push the right model to the right system on demand is what makes AI viable in these environments without requiring fleet-wide hardware upgrades.

Market framing

Slaughter frames the opportunity by comparison: Nvidia generates roughly $200B in annual revenue, while the Department of Defense spends around $1 trillion per year, with the Trump administration reportedly targeting $1.5 trillion. His argument is that defense tech is undersupplied relative to that capital base — implying not dozens but hundreds of defense unicorns should eventually exist in the ecosystem.