Andy Markoff

Markoff runs roughly 26 people who describe their shop as the first frontier AI lab built specifically for national security. The work is unglamorous in the way that matters: fuse multimodal sensor data, weigh the scenarios, hand a commander a recommendation across fires, logistics, intelligence, information, and force protection - and do it in the time a human has, which is usually no time at all.

The urgency is not a marketing register. It is a calendar. "Decision Dominance will be the deciding factor in preventing WWIII," Markoff has said, "because it's the only goal that's achievable before 2027." Most founders sell a roadmap. Markoff sells a deadline, and he treats every quarter as if the deadline were real.

What separates Smack from a deck full of buzzwords is where the founders learned the problem. Markoff and co-founder Clint Alanis are Marine Raiders, MARSOC veterans with more than two decades of combat experience between them. They did not read about coordinating distributed operations under fire. They did it, and they remember how long it took.

In March 2026 the company announced $32 million in combined Seed and Series A funding, led by Geodesic Capital and Costanoa Ventures, with Point72 Ventures, Felicis, Bloomberg Beta and others on the cap table. The money buys time against a clock Markoff refuses to pretend is further away than it is.

The resume reads like someone could not decide which thriller to star in. A B.A. in Political Science and International Relations from Princeton. A commission as a Marine Corps infantry officer. A Marine Raider tab. An operations officer's seat during the Battle of Mosul, against ISIS, where the gap between "we have a plan" and "the plan is already obsolete" is measured in lives.

Then the pivot most veterans never make: Palantir. Markoff joined the business operations and strategy team in Palo Alto, sitting between engineers and the messy reality of deployment, watching software meet the field and learning where it broke. He also passed through Origin Materials and ran Pallas Performance along the way.

The throughline is not the logos. It is the seat he kept choosing - the one where a decision has to be made now, with incomplete information, and where being right slowly is the same as being wrong. Smack is what happens when that person stops waiting for someone else to build the tool.

Geodesic Capital, when it explained its check, kept returning to one thing: the founders understood the coordination problem because they had bled over it. The harder part, as Markoff frames it, "is not only building AI reasoning models, but coupling that with domain expertise and experience" - and tuning it for a battlefield where bandwidth is scarce and the network may not be there at all.

Timing is a strange kind of luck, and Markoff caught a gust of it in the spring of 2026. The Pentagon deemed a much larger AI company's products a supply-chain risk, and overnight the calculus for small, security-native labs changed. The phones at Smack started ringing. U.S. Special Operations Command came calling. The Marine Corps, which had awarded Smack a contract in March 2025 and taken delivery of a working prototype by that October, suddenly wanted to talk about pulling production forward.

The originally planned timeline pointed at fiscal year 2027. Inside weeks of the shakeup, Markoff described "very specific guidance and movement and energy" toward getting the prototype combat-ready in 2026 - an acceleration of more than a year. For a founder whose whole pitch is about speed, the irony is almost too neat: the market finally moved as fast as he had been asking it to.

His answer was not to play it cool. "We want more, we want demos, let's talk about how we can move faster," he told a reporter, which is roughly the opposite of the careful hedging most enterprise founders deploy when a competitor stumbles. Smack now holds contracts across the Air Force, the Marine Corps, and the Navy, with two of them moving from prototype toward production.

The bet underneath all of it is narrow and specific. Not general-purpose intelligence. Not a chatbot for generals. Domain-specific models, fused with the lived experience of people who have actually planned operations under pressure, tuned to work when the bandwidth is thin and the stakes are not theoretical.

Markoff did not build Smack alone. Clint Alanis, his co-founder, is a fellow MARSOC veteran; the two of them carry more than twenty years of combat experience between them. The company also lists Dan Gould as CTO and co-founder, and has been hiring from the operational world it serves - bringing on a former Joint Fires Network chief engineer as a VP of Decision Dominance, a title that doubles as a mission statement.

That hiring pattern is the tell. A founder optimizing for a quick acquisition staffs differently than one trying to earn the trust of a skeptical Department of War. Smack keeps pulling in people who have stood in the rooms where these decisions get made, because the product only works if it speaks the language of those rooms.

The investor list reads the same way. Geodesic Capital and Costanoa Ventures led the round, joined by Point72 Ventures, Felicis, First In, Scribble Ventures, Bloomberg Beta, Washington Harbour Partners, Palumni VC, Fulcrum Venture Group, the Anomaly Fund, and Fortitude Ventures. Geodesic, explaining its check, called distributed operations at scale in contested environments "one of the hardest problems in defense" - and pointed to the team's training alongside Japanese forces as evidence the platform fits the Indo-Pacific, the theater everyone in this business is quietly planning around.

Strip away the funding noise and Smack is a wager on a single idea: that the next conflict will be decided less by who has the best hardware and more by who can orient and decide first. Markoff has lived the cost of being slow. He has built a company so that the people who come after him do not have to. Whether 2027 turns out to be the deadline he believes it is, the question he is asking - how fast can you decide, and can you decide faster than the other side - is not going away.