Scout AI builds foundation models and AI agents for unmanned defense platforms—ground, air, sea, and space. The tech stack is heavily weighted toward ML infrastructure (PyTorch, TensorFlow, JAX, DeepSpeed, Ray) paired with embedded systems (STM32, ESP32, ARM, RTOS) and orchestration (Kubernetes, Airflow), revealing a company bridging research-grade model training with hardware deployment constraints. Active projects span field trials, flight testing, and data pipelines for autonomy, while pain points cluster around deployment in adversarial environments and compute scaling—typical friction for moving frontier models from lab to operational systems.
Scout AI is a defense AI lab founded in 2024, based in Sunnyvale, California, with 11–50 employees. The company develops its Fury foundation model to transform unmanned platforms into intelligent agents for the U.S. Department of Defense and allied militaries. The product addresses autonomous systems, multimodal model fusion, and real-world field deployment. The team is engineering-heavy (9 active roles) with concentrated hiring in research and operations, signaling rapid scaling of both R&D and field validation infrastructure.
Fury, a foundation model that transforms unmanned ground, air, sea, and space platforms into AI-powered collaborative agents for defense and military applications.
Machine learning: Python, PyTorch, TensorFlow, JAX, DeepSpeed, Ray. Embedded: STM32, ESP32, ARM, RTOS. Infrastructure: Kubernetes, Airflow, AWS, GCP, Azure, Docker, Prometheus, Grafana.
Other companies in the same industry, closest in size