Money does not usually make noise. It moves quietly, through conference rooms and wire transfers, nodding at spreadsheets like they told a good joke. But every once in a while, capital clears its...
Money does not usually make noise. It moves quietly, through conference rooms and wire transfers, nodding at spreadsheets like they told a good joke. But every once in a while, capital clears its throat and says something worth hearing. Overland AI just did that with a $100M Series B plus venture debt round, and if you work anywhere near autonomy, defense, or the uncomfortable future where machines have to go where humans should not, you probably felt the bassline.
Founded in 2022 and built out of the University of Washington Robot Learning Lab, Overland AI came from a simple and stubborn truth Byron Boots, CEO, saw after nearly 10 years with the U.S. Army Research Laboratory. Ground autonomy is still brutally hard. Almost every military ground vehicle still has a human inside because dirt, mud, trees, snow, dust, and chaos do not care about your demo video. Byron Boots, Stephanie Bonk, President, and Greg Okopal, COO, decided that was unacceptable, then got to work making machines that can think while bouncing at speed through places GPS politely refuses to function.
The result is an autonomy stack that does not ask permission from the environment. OverDrive handles off-road navigation at tactically relevant speeds. OverWatch lets one operator manage multiple vehicles without white-knuckles. SPARK ties the hardware together. ULTRA carries 1,000 lbs, runs 100 miles, clears 14.5 inches of ground, and does it without waiting for someone to grab the wheel. Polaris RZR, Textron Ripsaw M5, General Dynamics S-MET. Same brain, different bodies, no babysitter required.
This is not lab theory. Overland AI was the only team to complete all 8 DARPA RACER field experiments and then took that autonomy into a live-fire exercise with the U.S. Army. Army units, Marine Corps units, SOCOM, and even CAL FIRE have already put these systems to work. When your tech survives mud, smoke, contested terrain, and real operators with real consequences, the conversation changes.
That is why 8VC doubled down again, leading both the Series A and Series B, with Joe Lonsdale backing execution over noise. Point72 Ventures stayed in, with Chris Morales on the board since seed. Valor Equity Partners and StepStone Group stepped in. TriplePoint Capital brought $20M in venture debt to keep the factory humming in South Seattle. $80M in equity, $20M in debt, $142M raised total, valuation unconfirmed, because some things do not need to be announced to be understood.
The lesson here is not about funding. It is about patience, field work, and earning trust one deployment at a time. Overland AI did not chase headlines. They chased terrain, contracts, operators, and repeatable performance. Manufacturing scaled. Integration teams grew. The tech moved from experiment to operational reality. Defense and dual-use customers do not buy vision decks. They buy outcomes.