Pirates Only

Adam Warmoth, founder and CEO of Chariot Defense, joins the Pirates Only pod to sit down with Andrew Couillard, who first met him two years ago when Chariot was still in stealth and being built out of a garage in South San Francisco. Since then, Chariot went from a seed check to air assaulting its systems into a force-on-force exercise with the 101st Airborne within six months, raised a Series A led by Andreessen Horowitz, and won a DIU contract to bring its Amphora power systems to an Army brigade. A Stanford engineering physics and mechanical engineering grad, Adam spent a decade on electric aircraft at Kitty Hawk, Uber Elevate and Archer before leading counter-drone engineering at Anduril.Adam explains why power is the missing layer of modern warfare after comms and compute, why units leave their generators behind as they get lighter and more distributed, and how Amphora works as a hybridization kit that takes power in from any source and puts clean, silent power out to drones, EW kits, radios and edge compute. He and Andrew dig into why the Army still plans in gallons instead of kilowatt hours, how a test of food, water and nicotine reveals real product market fit, why stakeholder support multiplies instead of adds, why the real competition is soldiers buying Chinese battery banks at Home Depot, and why he chose to be a solo founder in defense hardware.

Andy Markoff, co-founder and CEO of Smack Technologies, sits down with Andrew Couillard fourteen months after their first conversation. In that time, Smack has announced more than $90 million in fundraising, grown from a small team to a company winning contracts across nearly all services, and started building edge-optimized hardware that straps to a Marine's forearm. A former Marine Raider and special operations officer, Andy explains why large language models and monolithic cloud models fall short in a peer-level conflict, why most military decisions require geospatial and temporal data rather than text-based data, and how Smack's Omega platform for commanders passes campaign-informed reasoning and constraint harnesses to Alpha, its decision platform for small teams operating at the edge with no comms.The conversation covers what the Anthropic fallout with the Department of War meant for Smack's Series A week and for venture interest in domain-specific models, why Smack takes a hybrid, multi-algorithmic approach that includes reinforcement learning, operations research, and small language models, and how a culture of "and," an in-office fight gym, and a team where half the company was the end user shape hiring. Andy also shares how Smack competes for AI researchers against general-purpose frontier labs, why he believes the defense tech bubble leaves room for only a few new primes, why software-only defense companies need hardware to scale, and why he would create more joint program offices if he were king for a day.

Mike Berman, co-founder and CEO of Tectonic, joins Andrew Couillard to unpack the future of a post-quantum world. Tectonic entered the first Black Flag cohort in July 2025, and builds post-quantum cryptography for national security systems, including key management, crypto agility as a service, secure networking, and an identity and attestation layer, so the government can rip out RSA once and never have to rip and replace it again. Mike explains why a quantum computer in an adversary's hands would amount to a master key to everything we encrypt, from nuclear codes and power grids to every American's 401k, and why harvest now, decrypt later means the attack is already underway.The conversation walks through how Tectonic's government go to market stacked over time: a CRADA with NIST and a seat in the post-quantum consortium alongside Amazon, Google, IBM, and Palo Alto Networks, a Phase I Navy SBIR that became a contracting vehicle, advisors who served as CTO of SOCOM and CTO of Space Force, and the decision to spend half his time in Washington DC working with primes, lobbying on Capitol Hill, and meeting CTOs across the Department of War and the intelligence community. Mike, a former Eventbrite and Salesforce seller, shares how executive alignment and BANT translate from enterprise sales to the government, how CTO Mike Lodder and a team from NASA, Cloudflare, Microsoft, and Amazon get hired through relationships that have stood the test of time, and why he would cut facility clearance red tape if he were king for a day.

Simon Shuham, co-founder and CEO of American Supercritical, joins Andrew Couillard to explain why the exhaust heat pouring out of gas turbines may be the most overlooked power source for the data center boom. Based in Denver, American Supercritical builds supercritical CO2 power blocks that bolt onto a turbine's exhaust and turn waste heat into electricity, giving simple cycle turbines up to 50 percent more power from the same fuel in a tenth of the footprint of a steam system. The first product is a 10 megawatt skid aimed at the sub 50 megawatt, behind the meter generation that data centers are building to get around grid interconnection queues. Simon traces his path from rocket engines at ULA and Blue Origin, through Ursa Major, Harvard Business School, and Exowatt, and explains why a technology that lived in labs for decades finally has the right market, the right timing, and the right co-founder in CTO Matt Carlson, a veteran of Sandia National Labs and Kairos Power.The conversation widens into why steam has dominated power generation for two hundred years and why Simon believes supercritical CO2 is its next generation. Because the system is heat source agnostic, the same architecture from roughly 5 to 50 megawatts can pair with small modular reactors, thermal batteries, concentrated solar, and waste heat from steel and cement plants. Simon covers who actually buys the product (the developer owner operators deploying gas turbines), why incumbents like GE face a classic innovator's dilemma, what nine months and roughly fifty co-founder conversations taught him about patience, and how the team grew from two to seven in about a month by hiring deeply experienced engineers and building a flexible, family friendly culture. He closes with advice for future founders: be obsessed with an idea you have real expertise in, and never start a company just for the sake of starting one.

Sitreps began in 2019 as a military meme page Kyle Eberly started while doing M&A work at PwC, built on a simple insight: you do not need three million followers, you need a hyper concentrated audience of exactly the right people. Nearly every veteran in the professional world seems to follow it, and that reach became a recruiting flywheel once Kyle, Wyatt Frasier, and Max Cormier realized their community was the perfect talent pool for the defense tech surge. Their first real client, OneBrief, called and told them to point the platform at recruiting, and the crew listened to the market. Since then they have worked with dozens of companies and hundreds of placements, layering a serious business on top of the memes.From there the conversation widens into why this moment is different: defense tech in its current form is only a few years old, military experience finally commands tech salaries, and the hardest roles to fill are cleared technical talent and people who can sell novel technology to the government. Kyle frames the entire sector as a scalability problem, since no one can afford to fire a five million dollar missile at a five hundred dollar drone, which is exactly why venture money keeps pouring in. The founders argue defense tech is a frontier rather than a bubble, point to energy as the real constraint (one client is putting decommissioned naval reactors to work powering AI data centers), and tease what comes next for Sitreps: a pitch competition, community chapters, and possibly a fund of their own.

Sumant Sharma, co-founder and CEO of EraDrive, joins Mat to explain why the satellites circling Earth today are effectively blind and micromanaged from the ground, and why that model breaks the moment constellations scale into the thousands. EraDrive spun out of the Stanford Space Rendezvous Laboratory in 2025, but its core vision-based navigation software was already flying first, demonstrated on four NASA CubeSats as part of the Starling mission, the world's first instance of multiple satellites navigating relative to each other using vision alone. That reverse order, technology in orbit before incorporation, became the foundation for a $5.3 million seed round and five active customer contracts spanning civil, commercial, and defense work.Sumant lays out the three problems EraDrive solves: a coordination crisis as roughly 30,000 tracked objects head toward 60,000 or more within five years, a contested domain where adversary satellites jam and shadow US assets on a weekly basis, and commercial applications like orbital data centers and in-space manufacturing that simply cannot run with a human in the loop. He argues that autonomy in space is not optional the way it is for cars, and paints a future where a handful of engineers steer a million satellites by typing mission intent and walking away. Every captain eventually has to let the swarm steer itself. The conversation closes with hard-won fundraising advice for hard tech founders: sell milestones, not vibes, and let revenue and signed contracts do the talking.Episode Summary:EraDrive spun out of the Stanford Space Rendezvous Lab in 2025 after its software was already flying on four NASA satellites.The team is 12 people based in downtown Palo Alto, deliberately near Stanford talent rather than the larger hubs in LA and Colorado.Falling launch prices and cheaper AI tooling have collapsed the barrier to building space hardware, though go-to-market remains the hard part.Roughly 30,000 objects larger than ten centimeters are tracked in orbit today, with credible estimates of 60,000 to 70,000 within five years.Space traffic coordination still relies on the 18th Space Defense Squadron emailing operators when a collision is predicted.China, Russia, and India have demonstrated anti-satellite capabilities, and adversary jamming of US assets happens on a weekly basis.Gen 1 of EraDrive's product retrofits onto satellites already in orbit using off-the-shelf cameras and existing flight compute.Eighty to ninety percent of EraDrive's revenue today comes from government and defense primes rather than purely commercial operators.The long-term vision is a handful of engineers steering a million-satellite swarm through natural-language mission intent.Sumant's fundraising advice is to sell concrete milestones and de-risk with real customer contracts and revenue.

Amir Khan and Jon Haase came to Odin Dynamics from opposite ends of the same problem. Amir built batteries and powertrains at Tesla, ARC, and Matternet, then spent mid-2024 cold-emailing admirals, sailors, and special forces to learn where maritime warfare was headed. Jon spent nearly 25 years in the Navy as an EOD officer and program manager who owned most of the service's underwater vehicle portfolio, and his first reaction to Amir's pitch was to tell him to pack up and go home. What changed his mind was a Tesla Plaid battery pack, the kind of first-principles energy engineering that the underwater world had never seen. Together they are building a system that functions as both a UUV and a torpedo, unlocking a mission set neither category could reach alone.The conversation charts why the ocean may be a harder build than space, where water is a thousand times denser than air and a vehicle is effectively blind past a few feet. Amir lays out the Tesla playbook of low idiot index, deleting parts that add no value, and designing for a hull every 20 minutes rather than a beautiful one-off. Jon makes the strategic case that undersea remains the last invisible domain, a place where whoever masses smart, long-endurance platforms first earns a decisive advantage. They close on fundraising hard truths, the value of functional hardware over promises, and a shared belief that for any founder ready to chart a course into deep water, the sea is their oyster.

In this episode, Mat sits down with Dan, CEO of Firestorm, and Michael LaFramboise, CEO of Aurelius Systems, two founders tackling the drone revolution from opposite but complementary angles. Dan conceived Firestorm after watching ISIS use $3,000 DJI drones to fight the U.S. military to a standstill in Mosul, and spent the years since rethinking drone manufacturing from first principles: stop building artisan Ferraris and start stamping out Camrys. Firestorm now produces four drone models and its flagship XL product, a containerized expeditionary manufacturing facility that can be parked in any available lot and immediately begin printing drones at the point of need. Michael LaFramboise built Aurelius Systems around the other side of the same equation, bringing low-cost, AI-driven directed energy weapons to the counter-drone fight with the goal of collapsing the marginal cost of a drone shootdown to a dollar or less per kill, using electricity instead of missiles as ammunition.The conversation covers the full scope of where the drone arms race is heading and what it means for American defense. Dan lays out a stark production gap: Ukraine and Russia each plan to build three to seven million drones this year, while the U.S. will produce closer to 30 to 50 thousand. Both founders agree that the winners in drone and counter-drone are largely already decided, that new entrants face brutal supply chain and procurement headwinds, and that the real fight now is go-to-market rather than pure technology. They also look ahead to unmanned ground vehicles reshaping land warfare in Ukraine right now, the emerging threat of satellite blind-and-destroy operations in low Earth orbit, and the long arc toward directed energy systems in space. For any founder bold enough to chart a course into hard defense tech, the shared advice is blunt: hire a lobbyist first, demo constantly, and plan for everything to cost twice as much and take twice as long.

Eric McShane spent a decade in battery research before discovering, during his Stanford postdoc, that startups were the path he had been looking for all along. Together with co-founder Evan, he launched Electroflow in San Bruno to tackle one of the most consequential supply chain problems in modern energy: producing lithium iron phosphate, or LFP, the cheapest and most widely used battery electrode material on the planet. LFP goes into over half of all batteries produced today and 99% of it is manufactured in China. Electroflow's core innovation is a three-step process that produces LFP directly from lithium-containing brine waste streams, replacing the traditional 10-step method that relies on vast evaporation ponds and months of lead time. The target production cost is $3,000 per ton, half the current Chinese market price of $6,000.Fresh off a $10 million seed round and just over two years removed from a napkin sketch, Electroflow is weeks away from deploying its first full-scale electrochemical stack at a real brine site. The demand picture behind that milestone is staggering: Eric projects a 40X increase in global battery capacity between now and 2040, driven by grid and data center storage, electric vehicles, and what he calls the most under appreciated demand signal on the horizon: billions of humanoid robots. The episode goes beyond the technology to cover the equally demanding work of building the company, including how scientists-turned-founders need to rewire their pitch instincts, why over-secrecy about your idea harms more than it protects, and how the WHO hiring method helped Electroflow scale to 16 people without a single firing. The crew has barely lifted anchor.

Chandler Luzsicza spent nearly five years at SpaceX as a propulsion engineer on the Dragon capsule and Starship programs before a brief stop at autonomous vessel company Saronic convinced him to found Galadyne in late 2024. Backed by a pre-seed led by Andreessen Horowitz, his Austin team is applying the commercial space playbook to what he calls an American missile crisis. The numbers are sobering: congressional war gaming estimates suggest the U.S. has three to eight days of critical munitions available for a South China Sea conflict, and replenishing just 20% of THAAD inventory transferred to Israel during a recent conflict could take three to eight years. The cost picture is equally grim, with per-round prices on legacy platforms reaching $13 million before a standard two-shot doctrine doubles that figure.The root cause, Chandler argues, runs deeper than procurement dysfunction. Solid rocket motors depend on a single domestic ammonium perchlorate facility so hazardous that workers who handle it have historically been called "angel pushers," and standing up a new one would take a decade the country does not have. Galadyne's answer is to chart a new course entirely: liquid-propellant systems loaded at the point of use, so the production line handles only inert metal components and can scale toward gigafactory throughput, targeting tens of thousands of units within three to five years. The episode rounds out with hard-won advice for early-stage founders: build conviction from first principles before you pitch, and reach out to tier-one investors far sooner than feels comfortable, because the proof bar at large funds is often lower than founders assume.