Defense Disruptors: How Firestorm Turns Shipping Containers Into Frontline Drone Factories

The drones over Foxborough were not a national security problem. They were a football problem.

It was the summer of 2015, and Dan Magy was in the business of robotic camera systems for sports stadiums. He had been making the rounds, and at stadium after stadium the head of security had pulled him aside with the same question.

Did he know anybody building drone defense systems? People kept flying quadcopters over the venue. Some of them were flying over practice.

He had heard the question four or five times already and let it pass. At Gillette Stadium, home of the New England Patriots, he heard it again.

“They had people flying drones into their stadium and over their practice to probably see what Tom Brady was doing at the time,” Magy said, telling the story 11 years later from a factory floor in San Diego. “It’s like when my wife asks me to do something five times, I finally listen and do it.”

On the trip home from Boston he wrote the framework of a patent.

It is worth pausing on what else was happening that year, because the two stories converge and neither party knew it.

While American stadium managers worried about hobbyists spying on quarterbacks, Islamic State fighters in Iraq and Syria were working out what else a $2,000 quadcopter could do. By late 2016 they had the answer.

Gen. Raymond Thomas, then commander of US Special Operations Command, later described the result to an industry audience in Tampa as the most daunting problem his operators had faced that year: an adversary that, for a time, held tactical superiority in the airspace beneath American air superiority, using off the shelf aircraft and improvised munitions, with small arms fire the only available answer.

At one point, Thomas said, the Iraqi offensive at Mosul nearly stalled. Over 24 hours there were 70 drones in the air. At one moment a dozen of them were directly overhead.

The company Magy incorporated in 2016 out of that Boston drive was Citadel Defense.

Its product, an artificial intelligence driven radio frequency jammer called Titan, was designed to detect and force down small drones without requiring a signals intelligence specialist to operate it.

Magy said Naval Special Warfare found the company by searching the internet, tested the equipment, and that Citadel was the second such system into Mosul.

BlueHalo acquired Citadel in November 2021. Titan remains a program of record with US Special Operations Command, and the line now sits inside AeroVironment following its acquisition of BlueHalo.

Magy did not serve in the military. What the Mosul years gave him instead was a working education in what the equipment was for.

“I got to actually understand, specifically during the ISIS fight, that evil does exist in this world,” he said. “I would talk to guys who’d come back from villages in Iraq where you would have a room where they would murder everyone, a room where they’d cut everyone’s hands off, a sex slave room. It was atrocious. We think of the world we live in and how sheltered it is, and it really opened my eyes to how you are always going to, with humans, need to have some sort of defensive or offensive capability, unfortunately.”

He still hears from the other end of the transaction. About once a month, he said, someone tells him the system his team designed a decade ago saved their life. The most recent was a man in Yemen who said it stopped a wave of drones coming at him.

Six Weeks of Retirement

The exit was good. The retirement lasted about six weeks.

By early 2022 the war in Ukraine was forming, and Magy had been thinking about the economics of the ISIS drone campaign in a way that most people in the counter drone business had not.

He describes what he felt as respect. A non-state actor with a hobby shop budget had fought the world’s most expensive military to a standstill in the air for months, until the jammers arrived.

The first idea was narrow and arithmetical. Could anyone mass produce a $10,000 loitering munition? Because if $10,000 could take out a Russian tank, the Ukrainians had a war they could fight.

That is not the company that exists today, and the distance between the two is the subject of this story.

The founding team is the reason. Magy brought in Ian Muceus, whose background is advanced manufacturing and who holds more than a dozen patents in 3D printing, as chief technology officer. Muceus was the one who said to build the thing with printers.

The third co-founder, Chad McCoy, had spent more than two decades in Joint Special Operations Command and had personally used most of the systems on the market. McCoy pushed hard, and by Magy’s account, for modularity. On a mission you carried a drone to observe, a different drone to strike, another to collect signals. Why was there not one platform that took payloads?

Magy also brought a piece of institutional knowledge that has nothing to do with engineering. The first hire was a lobbyist.

“I’m really happy now that Trae Stephens at Anduril finally admits their first hire was a lobbyist,” he said. “Our first hire was a lobbyist as well. That’s just how the game is played.”

Stephens has said as much publicly, describing hiring a lobbyist in Anduril’s first week and meeting government officials three days later who became the company’s first customers. It is one of the least discussed facts about the current generation of defense startups and one of the most consequential.

Firestorm self funded for a year, got an aircraft flying, then went to raise. This was 2023, which does not feel like ancient history and, in venture capital terms, was a different climate entirely. The company was building a drone designed to explode.

“That precluded seventy percent of the investors from even looking at us,” Magy said.

A Drone With an API

The original product thesis was borrowed, deliberately, from consumer electronics.

“Think about it as a drone with an API,” Magy said. “You can put a variety of sensors on it. We wanted to lower the cost of ownership for this category and get the Department of War, or whoever in the world the customer is, away from the idea that you buy one drone for one mission. We think it’s stupid. You wouldn’t buy one iPhone for each app. You buy one iPhone, you put all the apps on it.”

The workhorse that came out of that thinking is the Tempest, a Group 2 aircraft with a 6-foot (1.8-meter) fuselage. Firestorm offers it with two propulsion options. The company has flown it in the South China Sea region and in Eastern Europe, in cold, humid, and hot conditions. 

Below the Tempest sits a small first-person view (FPV) quadcopter, the Squall, co-developed with the Croatian manufacturer Orqa and announced in February.

Orqa’s FPV hardware has been used in Ukraine since the early phase of the war, and its principal selling point to American buyers is what it does not contain. The Squall is built in the United States from components compliant with the National Defense Authorization Act, meaning no Chinese parts.

The commercial logic there is not subtle. China controls an estimated 80 percent or more of the global drone component supply chain, and Beijing has been tightening the tap. On 5 August 2026, China’s Ministry of Commerce announced that exports of controlled drones and related items bound for the United States would face strict case by case review with no license facilitation.

What Magy said customers respond to is not the airframe. It is the swap.

“The number one piece of feedback on the Tempest is that the end user, because of how easy it is to put different payloads on it, is using it for a number of different things,” he said.

“You can do jamming, you can do strike, you can do collection. That’s the beauty. You have a trade space, and if something is in this weight, you can put it in there, and we can be integrating and flying it in hours or days, not weeks or months, which is the norm in this space, unfortunately.”

The same logic governs the radios, and here Magy’s first career is doing the talking. Having built a radio frequency counter-drone system, he designed Firestorm’s aircraft on the assumption that its communications would be defeated eventually.

“We build the truck, put different comms loads on it,” he said. “If there’s a radio that’s working really well for a number of months or a year, and then it doesn’t work, great. Put a new radio on it. We’re putting new radios on the drone in less than a day, for the most part.”

The Pivot Nobody Planned

About two years ago, the requests started arriving from an unexpected direction.

Firestorm’s customers had noticed the industrial 3D printers. What they asked for was not a better drone. It was the factory that made it.

“We started getting asked by customers, could you actually take your industrial 3D printers, put them in a shipping container, and then send us all the supplies so we can build drones where we need them to be?”

The answer became the xCell, an expeditionary manufacturing cell that today consists of two 20-foot (6-meter) shipping containers, one for printing and cleaning, and one for post-production assembly and R&D, running HP Multi Jet Fusion 3-D Printers under a partnership with HP. A loaded xCell fits on a C-130. Firestorm says it can deploy the unit and produce combat ready drones in under 24 hours.

Then came the part nobody had modeled.

“What we’ve found, just kind of on the side, is that when you give the end users industrial-grade additive manufacturing, they have a lot of problems they have to fix in the field that are not related to drones whatsoever,” Magy said. “Almost half of our prints now are actually printing parts to help solve supply chain issues.”

He locates the realization at a specific place and time. The first xCell ever built went to an exercise in the California desert. Two weeks, 120 degree heat.

“We fixed like 20 other people’s drones. We fixed a ton of vehicle parts,” he said. “That was an aha moment, because so many people, once they discovered that they had an industrial 3D printer in the field they could use, they could take deadlined systems, and bring them back to life.”

Deadlined is the term of art for equipment that does not work. It is also, quietly, an enormous market.

The United States operates a vast inventory of legacy platforms whose original manufacturers have moved on, whose tooling has been scrapped, and whose spare parts arrive on a schedule set by a supply chain designed for uncontested logistics.

The Department of War has made contested logistics one of six critical technology areas identified by the Office of the Under Secretary for Research and Engineering, and the reason is geographic.

In a Pacific conflict, the nearest American drone factory is thousands of miles from the fight, and the ships and aircraft carrying parts to the front are themselves targets.

Magy’s position is that the sustainment work is not a distraction from the drone business. It is the moat.

“There’s a way for you to be sticky, and to tie today’s requirements and future requirements together around a manufacturing play, but also around an autonomous play,” he said. “That stuff’s not gonna go away.”

Why Not Just Buy a Mold

Firestorm sells injection molding too. Magy is careful to say so, and careful about where printing stops making sense.

“Molds are very expensive,” he said. “If you look at traditional manufacturing, it’s very good for high throughput, so we offer molds as well, not just 3D printing.”

The case for printing is not cost per unit. It is the cost of being wrong.

He offers a live example. The industry is migrating from radio frequency links to Starlink, and quickly. Imagine, he said, that you did not design your $300,000, $400,000, or $500,000 molds with a Starlink terminal in mind. Now you go back and build a new mold before you can even begin testing.

“For us, we just 3D print a part, and the next day we can be testing, putting a Starlink on our drone, as opposed to having to go back and start from scratch on a design.”

The field modification requests, he said, are the single most consistent thing he hears from the US military and allied forces, and often they have nothing to do with Firestorm’s own aircraft.

“Sometimes it’s, hey, we found this problem on somebody else’s drone, we want to print a part. We test it on one, and then they’ll ask us to print 1,000 of these parts to go modify all the other drones.”

On capacity, Magy gave numbers rather than adjectives. Per printer, per month, whether the printer sits in the San Diego factory or in a container downrange: dozens of large Group 2 aircraft, or hundreds of FPV drones. Capacity scales by adding printers.

The materials argument is where he is most pointed, and it lands on an industry orthodoxy.

“Traditional manufacturing for drones is everything’s carbon fiber,” he said. “Well, for a drone you’re using one time, why are you doing that? If you only need to have something fly for three, four, five, six, eight hours, you can build things that are substantially cheaper and easier to put together, and also not reliant on Shenzhen for that carbon fiber. That’s something a lot of people don’t really talk about, is how much of our industrial base is relying on the Chinese supply chain, even for the basic products.”

The powder Firestorm prints with today is PA12 nylon, produced by suppliers in a number of countries outside that chain.

A December 2025 analysis by the Center for Strategic and International Studies identified carbon fiber, rare earth magnets, lithium ion cells, and gallium nitride chips as the critical chokepoints underpinning modern drone warfare, and noted that aerospace grade carbon fiber capacity sits with a handful of firms and cannot be surged.

Whatever else is true about a nylon airframe, it does not queue behind an autoclave.

The War That Changed the Conversation

Ask Magy whether Firestorm is a drone company or a manufacturing company and he will tell you how he raised his last round, and then tell you why the question got harder.

“I raised my last round of money positioning it as a manufacturing layer,” he said. “But what we’ve seen with how many drones were used, and used successfully, or unsuccessfully, in the conflict with Iran, the demand for our drone is through the roof. I think doing both together is going to be where we’re going to continue to play, because they are very symbiotic.”

The conflict he means began on 28 February 2026, when American and Israeli forces launched close to 900 strikes in 12 hours under the US code name Operation Epic Fury.

Iran’s response was, above all, an aerial mass campaign.

The United Arab Emirates alone reported intercepting more than 2,200 drones and over 500 ballistic missiles by early April. Saudi Arabia was targeted with at least 435 drones. Qatar took 87 drones alongside more than 200 missiles. Jordan, Oman, Kuwait, and British facilities in Cyprus and the Indian Ocean were all struck.

American losses in the same period reframed a long running argument about what survives contested airspace. It was reported in May that Iranian forces destroyed more than two dozen MQ-9 Reapers, close to a billion dollars of aircraft and roughly 20 percent of the Pentagon’s prewar inventory of the type.

Air Force Chief of Staff Gen. Kenneth Wilsbach nonetheless told the House Armed Services Committee on May 20 that the Reaper had been perhaps the most valuable player of the campaign, and that no other platform was close. Both things were true at once, which is the uncomfortable part.

For Magy the significance was institutional. It changed what he heard inside the building.

“The conversations we had in the Pentagon, I would say up until the Iranian war, they were like, drones are a Ukraine thing. I would hear that all the time. And that’s crazy to hear. But that was normal, and now you’re seeing drones wreak havoc across all of the Middle East, and the cat’s totally out of the bag.”

His broader reading is blunt and not especially comfortable for anyone who spent the last 20 years planning around American air dominance.

“For a long time, air superiority was a given for the entire West. It’s not a given anymore. We no longer have a monopoly on precision guided strike either, and so you have to basically adapt to the changing environment.”

The interest in a factory you can move, he said, is now heaviest from the region that just spent six weeks under drone attack.

“It’s now a war of infrastructure. How do you actually survive? How do you make things to stay in the fight?”

The Televised War

There is a second, quieter thesis Magy holds, and it concerns information rather than hardware.

Firestorm has been active on LinkedIn and built a large following doing it. Magy’s explanation is that the material is mostly already public. What is missing is somebody willing to gather it and present it so that people can absorb it.

“If you can simplify, and with visuals, show the story, that’s what we found,” he said. “And if you watch all the stuff, you can start to become an expert as well.”

He traces the pattern to the Global War on Terror, when ISIS posted video continuously, and to the Syrian campaigns that followed, when it became possible to watch a battle unfold day by day from a distance.

“We never really have slowed down since then,” he said. “The war is actually televised now, but it’s all through FPV and cell phone. It’s amazing.”

The observation is not incidental to his business. It describes the environment in which his customers, his competitors, and his adversaries now learn. Techniques propagate in days. A modification that works in Donetsk on Monday is being copied elsewhere by Friday.

That tempo is the strongest argument Firestorm has for building the way it builds, and it is also the reason Magy expects the field to be crowded.

The Money, and What It Buys

Firestorm closed an $82 million Series B on 29 April 2026, led by Washington Harbour Partners, with participation from NEA, Ondas, In-Q-Tel, Lockheed Martin Ventures, Booz Allen Ventures, Geodesic, Motley Fool Ventures, and Litquidity Ventures.

NEA had led the $47 million Series A the previous July.

Total funding to date is $153 million. Headcount went from about 40 to more than 160 in 12 months.

Eight days after the round, on May 7, the company announced an award of up to $30 million through the Department of War’s Accelerate the Procurement and Fielding of Innovative Technologies (APFIT) program, expandable to $50 million.

APFIT does not fund research. It funds buying and fielding things that already work. The package covers five xCell units, more than 200 Tempest aircraft, and operator training for an undisclosed Indo-Pacific customer. Roughly $26 million has been obligated across five task orders, and deliveries are under way.

These are Firestorm’s first deployments outside the continental United States. The company also holds a $100-million ceiling indefinite delivery, indefinite quantity contract with the Air Force, a STRATFI award, and an $18 million ceiling SBIR Phase II for the xCell.

Asked what the money actually buys, Magy started with the least glamorous answer.

“In the short term, we have to execute on the hundreds of millions of dollars of contracts we’ve won,” he said. “We’ve started to deliver xCells and drones this year. We have orders in the hundreds for drones for the bigger guys.”

Beyond delivery, three things. Overseas markets, with the strongest traction in Asia and the Middle East and enough of it to justify putting people on the ground.

Research and development to expand what an xCell can produce, since PA12 nylon is the only material it runs today and the government has been asking for more. And new platforms, some of them requested by name.

“We have the trust from some of our customers, so they’re now telling us what they would like us to build next, which is great.”

On Europe, he was less diplomatic, and the complaint is one heard increasingly from founders on both sides of the Atlantic.

“It’s going to be incredibly important for the Europeans to actually spend money. That’s one of the things I talk to all my startup friends in Europe about. A lot of talk, not a lot of contracts are coming.”

The data does not entirely contradict him. Venture funding into European defense startups reached 7.4 billion euros ($8.6 billion) in 2025, and the European Union’s 150 billion euro ($174 billion) SAFE loan program now ties major contracts to building at scale with components sourced inside the bloc.

But analysts and founders continue to report that the money is landing with incumbent primes rather than new entrants, partly because Europe is spending most of it filling urgent capability gaps rather than buying future warfare technology, and partly because short contracts and political reversibility make large capital expenditure irrational for anyone.

France is the partial counterexample, with the share of defense spending reaching small and mid sized firms rising from 14 percent in 2022 to 25 percent in 2024 after procurement reforms.

Magy’s warning to his own industry follows directly from the gap.

“If we’re going to be serious as the West collectively in being able to build the amount of drones we’re gonna need to fight and successfully win wars, or deter wars, in the future, the money needs to continue to flow into the unmanned space,” he said.

“We can help tackle some of that by tying this to the repair parts stuff we’re doing, because it’s so important for legacy systems to stay in the fight. But if you’re a one trick pony, governments better be spending money, or there’s gonna be a lot of drone companies that go out of business in the next couple of years, frankly.”

‘This Will Probably End Up Being a Bubble’

Magy used the word without being asked to.

The numbers around him are extraordinary. Venture investors deployed a record $19.8 billion into defense technology across 262 deals in the first quarter of 2026 alone, according to PitchBook, against $5.7 billion in the same quarter of 2024.

Crunchbase counted more than $14.6 billion in the sector through the first five months of the year, already past the $9.6 billion recorded for all of 2025.

Anduril, the company that was described in 2018 as the most controversial startup in technology, is now valued at $61 billion.

Early stage defense companies are raising at revenue multiples that run from 17 times to 50 times and occasionally higher.

Magy’s account of why is less about patriotism than about portfolio mechanics.

“It was impossible to raise money 10 years ago,” he said. “ESG standards, defense is tough. And I think there’s this perfect storm right now where they see warfare is changing, there are some of the right signals coming out of the Department, and out of some of the European countries. And you also gotta just appreciate that SaaS is in trouble. So you have a lot of VCs that have to deploy money.”

What worries him is not the capital. It is what the capital is buying.

“My concern is a lot of people coming in here don’t actually understand or even bother to call the customer. It’s kind of crazy to me. The customer will tell you where the roadmap is.”

He named four things he believes will carry Firestorm through a correction, and none of them are technology.

The first is proximity to the customer, which for him is a scheduling commitment rather than a slogan. He goes to Capitol Hill once a month, personally, and the company now keeps an office in Washington.

“Having a presence in DC is really important, because you have both the Pentagon and then the people who actually write the budget there.”

The second is the modular ethos, which he frames as an insurance policy against being wrong about the future.

The third is the repair parts business, which ties the company to systems that will still be in inventory long after the current drone cycle turns over.

The fourth is a deliberate refusal to build everything in house.

“We just admit to everyone, we can’t do everything on our drone in house, so we end up partnering with tons of different companies, and you have a little bit of a platform effect,” he said.

“A lot of those products, if it’s a camera, or an algorithm, or a sensor, they don’t have something to put it on, and that’s where the Tempest has really played a major role in allowing lots of different people to get their products into the end user’s hands.”

Then he named the thing he thinks will actually kill companies, and it is not a technology risk or a market risk. It is a delivery risk.

“Building something that works, and being able to build it pretty much on time, that’s ninety percent of the battle,” he said. “People forget how hard it is to actually build stuff, and I think a lot of rounds are being raised without understanding that meeting delivery schedules is really, really, really tough. It’s easy to tell a story and go to a demo. It’s a lot harder to mass produce something.”

He also acknowledged the obvious counterpressure on a company that just raised $82 million and quadrupled its headcount.

“A lot of that will come down to leveraging new technologies to make sure we don’t get too big. We are doing a lot, so we have to deliver on our contracts.”

Three Aha Moments, and the Third One Is People

Asked to identify the moments when he knew Firestorm was working, Magy named three, and the order is instructive.

The first came roughly three years ago, on an early flight with the Air Force Research Laboratory. The team performed a payload swap in front of the customer.

“They go, holy cow, that takes us six months with everybody else, you guys did it in an afternoon.”

The second was the California desert, the 120 degree heat, and the 20 other people’s drones.

The third has nothing to do with hardware.

“It’s really about people,” he said. “We’ve had some of the best in class people in the unmanned space seek us out to join us over the last year.”

A number of them came from General Atomics, which is a slightly awkward thing to say out loud in the year that Iranian air defenses removed roughly 20 percent of the Reaper fleet. Magy said it anyway, and defended the company he is recruiting from.

“Say what you will about how many of their drones get shot down. Without General Atomics, you don’t have Shield AI, you don’t have Kratos, you don’t have Anduril. They all come from GA.”

He said a group of senior, creative people had left recently to work on what he described only as a special project.

“We’ve had people leave vice president roles at other drone companies to come work here to become individual contributors, because they believe in the mission. That’s really, really cool. It’s when the end users tell you this is special, and you know who they are, so it means more. And then when the people are willing to leave jobs.”

What It Takes

The last question of the interview was the one about advice, and Magy declined to make it about defense.

“I won’t say for defense tech, because I think it’s just getting more and more convoluted,” he said. “I just think, in general, if you want to be an entrepreneur, the number one thing is to make the decision to go do it.”

He allowed that there are two legitimate schools. Keep the job and build on weekends until it works. Or jump.

He jumped, and he thinks he had an unfair advantage doing it, though not the kind that shows up on a resume.

“I have a superpower that other people don’t have. I’m incredibly comfortable in uncomfortable situations. I’m not exaggerating, I could live out of a cardboard box. I don’t need a lot of creature comforts, and I think that’s what gave me a really unfair advantage as a young entrepreneur. I didn’t need the nice car, I didn’t need the nice apartment. I don’t have any of that stuff now.”

The second lesson is the one that separates early stage from later stage, and he framed it as a warning.

“At the later stages, it’s to understand that this is not your baby, and you have to bring in people to help you scale the business. You have to trust people. You have to admit at times there’s things that other people are better at than you, because if you don’t do that, and you keep things too close to the chest, you’ll never be able to build a big organization.”

Three lessons, then. Jump in. Be comfortable being uncomfortable. And never treat the business like it is your child.

Toward the end of the conversation the discussion drifted into sport. The observation on the table was that the best football happens in the last two minutes, when everyone is exhausted and somebody is losing.

“Everyone’s tired,” Magy said, and then reached for his own example. “That’s why Argentina was so fun to watch in the World Cup. They scored all their goals after the 90th minute. It was incredible.”

It is a fitting note for a company built on the premise that interesting things happen after the plan stops working.

Firestorm’s entire product argument is that the airframe you designed will need a part you did not anticipate, that the radio will be jammed, that the factory will be too far away or will be hit, and that the answer is not a better forecast but a shorter distance between the problem and the machine that solves it.

Whether the market agrees for long enough to matter is the open question, and Magy is not pretending otherwise. He expects a bubble. He expects companies to fail. He thinks the ones that survive will be the ones that shipped on schedule and stayed close enough to their customers to know what to ship next.

In the meantime, roughly half of what comes out of his printers in the field is not a drone at all. It is a bracket, a housing, a fitting for a vehicle nobody makes parts for anymore.

That is not the business he set out to build. It may turn out to be the more durable one.