The US Army is turning to Austin-based Striveworks to put battlefield AI to the test, aiming to ensure commanders can trust machine learning tools when it matters most.
Rather than developing a single new algorithm, the effort centers on how AI models are tested, compared, and validated before they are deployed — a challenge that grows as the number of AI-enabled tools in the force expands.
Striveworks’ AI operations platform is designed to standardize how models are evaluated, giving commanders clearer insight into performance, reliability, and risk across different mission contexts.
The company says its system enables real-time comparisons between models, analyzes metadata to surface hidden vulnerabilities, and applies large language model-based testing to accelerate decision-making.
Those capabilities are aimed at ensuring AI tools can be deployed at scale without introducing security gaps, performance blind spots, or costly delays in operational environments.
Preparing for AI-Dense Battlespace
Striveworks’ test and evaluation services can operate independently or as part of its broader AIOps platform, which is already supporting the service’s Next Generation Command and Control (NGC2) program.
The army’s interest reflects a wider shift toward managing AI as a dynamic ecosystem, rather than treating individual models as static software tools.
“Tomorrow’s battles will involve hundreds — if not thousands — of AI and machine learning models, each with distinct capabilities and vulnerabilities,” said Eric Korman, Striveworks’ chief science officer and co-founder.
“In that complexity, hesitation or guesswork is costly,” he added, noting that commanders need objective data to make rapid choices under pressure.
He further stated that Striveworks’ approach helps leaders “instantly select and deploy the right model for the mission,” turning AI testing from a back-end technical process into a frontline enabler.
Striveworks earned a Phase II Small Business Innovation Research award following its performance at the army-backed xTechScalable AI 2 competition, part of a broader effort to accelerate trusted AI adoption across the force.
