Military logistics is not a back-office detail. For the U.S. Department of Defense, it is a central operating problem: how to move people, equipment and supplies through contested or disrupted conditions while staying within real-world limits on time, resources and budget.
Defcon AI, a Virginia-based startup incubated two years ago, is building software for that problem. The company has raised a $44 million seed round led by Bessemer Venture Partners, with participation from Fifth Growth Fund and Red Cell Partners, among others.
A planning problem with many moving parts
The Department of Defense is described in the source article as a mammoth organization. It employs millions of service members and hundreds of thousands of civilian employees, and it operates with the world’s largest military budget.
That scale creates a coordination challenge. Defense operators must decide how to plan operations, coordinate resources and remain within budget even when events are contested. Those events can include a hurricane or an adversary.
Air Mobility Command, a command of the U.S. Air Force, shows why the problem is so difficult. When operators plan airlifts, they must account for available aircraft, required crews, crew rest locations, refueling points, relevant airfields and cargo handling locations.
Those choices are connected. A change in one area can affect cost, timing and resource needs elsewhere. Defcon AI’s pitch is that software can help operators manage that complexity faster than manual planning alone.
How Defcon AI wants the software to work
Defcon AI says it has built software that lets an operator set planning parameters at the front end and then allow the system to generate a plan. Paul Selva, Defcon’s co-founder, chief strategy officer and retired U.S. Air Force General, told TechCrunch that operators can set the inputs “and then turn the software loose.”
The output is meant to include the best plan under the chosen constraints, along with cost tables, resource requirements and a schedule. In plain terms, the software is not just presenting a map or a checklist. It is intended to turn operational limits into a usable logistics plan.
The company uses a mix of algorithms, including machine learning and mathematical optimization algorithms. According to the source, the system can simulate a given scenario and generate the best logistical outcome for it.
Defcon AI initially used reinforcement learning algorithms that do not require data. The company now says it is ingesting more and more data provided by the DOD to power the software.
Operators can also decide whether the system should simulate how an adversary might disrupt operations. They can ask the software to optimize for different variables, including speed versus cost effectiveness.
Why the founders know the customer
Defcon AI was co-founded in 2022 by Selva, Yisroel Brumer and Grant Verstandig. Brumer and Verstandig are founding partners of Red Cell Partners, and Verstandig is also CEO.
Selva’s background is directly tied to the planning problems Defcon AI is targeting. During his military career, he served as commander of the Air Mobility Command, which oversees nearly all of the Air Force’s fleet of air lift aircraft. He later became commander of the U.S. Transportation Command, which coordinates transportation missions around the world, including those delivered by ships, trucks, trains and other forms of transportation.
Before retiring in 2019, Selva was nominated by President Barack Obama to be the vice chairman of the Joint Chiefs of Staff. His view of the product comes from work close to the operational roles the software is meant to support. “I’ve had all the jobs that we’re actually impacting,” Selva said.
Brumer’s experience also sits close to defense planning. Before joining Red Cell, he worked at the Pentagon as acting director of OSD/CAPE, the Office of the Secretary of Defense, Cost Assessment and Program Evaluation. He described that role as essentially the “chief analytics officer” for the DOD and the overseer for the budget submission process.
Verstandig has built or incubated companies including Rally Health and defense startup Epirus. Red Cell Partners, where Brumer and Verstandig are founding partners, makes internal investments and also incubates companies, including Defcon AI.
From mission need to real-world deployment
Selva said Defcon AI began because Air Mobility Command identified a mission need that industry was not filling. The founding team considered whether the problem could be handled through math and software. Their conclusion was that it was tractable, but required a different approach.
Brumer described the target as a problem of “maximal complexity.” That phrase fits the nature of defense logistics as described in the source: many resources, many constraints and little time during a crisis.
The company has earned around $15 million in government contracts. It has also delivered a production version that was deployed for a real-world operation with Air Mobility Command less than two years after founding.
Defcon AI is now working to certify the software to handle classified, secret information. That effort is meant to expand its uses inside the DOD and allow it to ingest even more data.
The company is also expanding its planning and simulation software beyond air operations. Its roadmap includes trucks, trains and ships, which would move the product closer to the broader transportation missions Selva previously oversaw.
What comes next for Defcon AI
Defcon AI does not appear to be treating Air Mobility Command as the endpoint. The company sees more applications across the DOD where its software could make an operational difference.
Brumer also said the company is seeing “a very strong demand signal” from the private sector. The source does not detail those private-sector uses, but the basic appeal is clear: organizations that must move resources under constraints may face planning problems that resemble parts of defense logistics.
Still, Defcon AI’s core claim is strongest where the end users are deeply involved. The company says working closely with operators can lead to a better product and a competitive edge in adversarial situations.
Selva framed the challenge around risk. “Operational planners are actually trying to assess risk for their commanders,” he said. “They’re probably the most skeptical audience for decision support tools, so the extent to which you can partner with them you achieve a better outcome.”
That skepticism matters. In military logistics, decision support software has to be useful under pressure, clear about tradeoffs and connected to real operational constraints. Defcon AI’s $44 million seed round is a bet that this kind of planning can be improved with software built around the way defense operators actually work.