Award

The Army’s strategic environment demands persistent deep sensing in contested and dynamic settings. While the military Internet-of-Things (MIoT) offers a framework for intelligence gathering, its reliance on centralized processing and high-bandwidth communication poses significant challenges in austere environments where connectivity may be limited or denied. Ozni AI proposes NEMESIS (Networked Edge-based Multi-INT Exploitation for Situational Intelligence and Security), a novel AI system that addresses these challenges by federating sense-making capabilities directly onto sensor platforms. NEMESIS leverages new AI techniques, including foundational Large Language Models (LLMs) and multi-INT embeddings (EO/IR, RF, audio), to enable sensors to interpret and understand the context of their local environment in order to estimate the pattern of life and detect anomalies. This processes enables the generation of trusted, traceable intelligence in real-time at the tactical edge, and provides actionable insights to warfighters and commanders.