He wanted the software so badly he had to build it himself. Now his machines watch the battlefield so people don't have to.
CEO & Co-Founder, TurbineOne — San Francisco
Ian Kalin: politics major, reactor engineer, and the guy who reads a battlefield the way you skim a headline. // Photo: TurbineOne
A founder usually pitches you the future. Ian Kalin pitches you a memory. "This is the product I wish I had when I served in the Navy," he says about TurbineOne's Frontline Perception System - and you believe him, because he spent five years on the other end of the problem, staring at more sensor feeds than any human was built to process.
TurbineOne builds one strange, specific thing: machine learning that runs on the sensor itself. Not in a datacenter. Not in the cloud. On the drone, the camera, the device sitting in a place with no signal and no patience. Kalin's rule for it is blunt - treat battlefield AI like a self-driving car, not a chatbot. "Intelligence has to live where the data is generated," he says. A chatbot can wait for the network. A soldier can't.
The company calls it the Frontline Perception System, and it is aimed squarely at a problem Kalin knows in his bones. Modern militaries drown in imagery. Cameras, radar, drones, satellites - all of it pouring in faster than eyes can review. The old fix was to add more eyes. Kalin's fix is to teach the sensor what matters and let it flag the rest.
Ask him who his real competitor is and he doesn't name a company. "Our biggest competition is manual labor," he says. "If you have enough bodies to throw at more screens, and you can just keep throwing 25-year-olds to zoom and scroll all day, you don't need me." He wants to retire that job.
Kalin and co-founder Matt Amacker met at The General Partnership, a venture firm whose anchor backer is Reid Hoffman. Amacker had prototyped at Amazon, Google, and Toyota Research. Kalin had a resume that reads like three different people stitched together.
Before a single line of code, they wrote the company's ethics policy - integrity, responsibility, accountability, non-partisanship, prosperity. In defense tech, where the ethical questions arrive on day one, Kalin decided to answer them on day zero.
The bet is working. In May 2025 the company closed a $36M Series B led by The General Partnership, with Bessemer, XYZ, Stepstone, PROOF and Artisanal along for the ride, at a reported $300M valuation. Already working with the Army, Air Force, Navy and Marine Corps across the Indo-Pacific, TurbineOne is now angling toward the intelligence community and a footprint overseas.
"We're helping the good guys find the bad guys."
// Ian Kalin, on what TurbineOne actually does
A School of Foreign Service degree - the least likely on-ramp to a reactor.
Counter-terrorism officer, then nuclear engineer aboard the carrier USS Ronald Reagan.
Led the U.S. Energy Data Initiative under U.S. CTO Todd Park.
The department's first-ever Chief Data Officer, across all 12 bureaus.
The product he wished he'd had, finally built.
Reported $300M valuation; eyes on the intelligence community.
Bars scaled for comparison, not to a single axis.
"We must treat battlefield AI more like a self-driving car than a chatbot."
On physical AI
"Our biggest competition is manual labor."
On the real rival
"This is the product I wish I had when I served in the Navy."
On why he built it
Deliver the best mission-AI to the nation's frontlines - real-time machine perception on the sensor at the tactical edge - and push TurbineOne past the armed services into the intelligence community and allied nations abroad.
The through-line across five careers is the same instinct: unlock the data that's already there, and put it in the hands of the person who needs it, right now, wherever they are.