Military AI is going on a little Eurotrip, y’all.
This morning, AI-for-defense hotshot Striveworks announced in an exclusive release to Tectonic that it’s scored a $9.75M contract with the US Army in support of Task Force RAVEN, the “Army effort to put AI agents to work in European operations.”
Under the contract, US soldiers in Europe will be able to use Striveworks’ Chariot Forge software to “carry out complex, time-sensitive tasks, including intelligence assessments, target validation and fire-support coordination, with human review and authorization built into the process.”
- Task Force Raven is a US Army unit in Europe that tests and rapidly integrates emerging technologies—especially drones, autonomy, and counter-UAS systems—into real-world operations.
“There is a lot going on in Europe right now,” Striveworks CEO Jim Rebesco told Tectonic. “It’s important to be able to kind of use this stuff in dynamic environments because that’s going to be where you build that human-machine trust. That means if there ever is active conflict, you can actually use these [AI] systems.”
“Task Force RAVEN is setting the pace in bringing agentic AI into intelligence assessments,” he added in a statement. “[Chariot Forge will give] the team the ability to continuously test, evaluate and apply guardrails to agentic workflows with the rigor this mission demands.”
The new award comes less than six months after the company scored a $70M multi-year enterprise agreement (EA) through the US Army to deploy their software across the Pentagon—this new award is against that EA.
Back it up: We’ve covered Striveworks before, but here’s a bit of a refresher for those less familiar with the AI-nuts-and-bolts of defense.
The company was founded back in 2018 by Rebesco, Craig Desjardins, Eric Korman, and Tony Manganiello in Austin, TX, around a pretty simple idea—they wanted to bring all of the capabilities we civilians have access to every day through things like ChatGPT to the military.
That’s (put very simply) what Chariot Forge does.
- Think of Chariot as the plumbing between an AI model, all the military’s messy and sometimes classified data, and the person who actually needs an answer.
- Users can plug in different models, connect them to their own data, build AI-powered workflows, and then deploy those workflows from the cloud all the way out to disconnected or compute-constrained environments at the edge.
- Critically—Striveworks isn’t trying to build a giant model that does everything. The platform is designed to be model-agnostic—meaning customers can swap in new or better models without rebuilding the whole system around them.
- And it adds the decidedly less sexy—but pretty important—stuff around those models: testing, monitoring, traceability, deployment, and controls over what the AI is allowed to do over the course of its lifetime.
- All this means that a soldier or analyst can ask a question in plain English, have the system pull from the relevant military data, and get an AI-generated answer back with the underlying source material attached so they can check the work.
- Per the company, Chariot Forge helps users process “intelligence 40 times faster and targets 21 times faster [and that] recent testing and evaluation improved targeting precision by nearly 45 percent.”
And this model has proven pretty popular with the US military—in addition to the Army, Striveworks says it has demoed this capability with SOCOM, the Navy, and across EUCOM, INDOPACOM, and CENTCOM.
Smarty pants: Under this Task Force RAVEN contract, a big focus will be on the intelligence analysis side of things—Rebesco sees that (as well as the great leap across the pond) as pretty major for his company.
“It’s a huge deal to be able to get into this new domain—intelligence and intelligence assessments, and to start delivering that continuous improvement and continuous quality control of them,” he said. “[Plus] we’re looking to grow our presence in Europe significantly, not just with…U.S. forces, but with NATO and with partner nations as well.”
