Tactical Automatic Speech and Entity Recognition System (TASER)

In recent years, there has been exponentially more amounts of unstructured natural language data from across the battlefield such as radio transmissions, audio, chat, and text.Ā Currently, this data needs to be received, acknowledged, analyzed, and understood manually by warfighters.Ā This process is often slow, error prone, and not scalable since a human is needed on both ends of every channel of communication.Ā As the size, weight, power, and cost of high-performance AI chips has fallen, a massive opportunity exists to create a resilient and scalable edge-based natural language processing system to intercept these data streams, clean and label them, and pass onto other software systems to automate battlefield events.Ā With this technology gap, OKSI proposes the Tactical Automatic Speech and Entity Recognition System (TASER) to provide resilient automated battlefield natural language processing at the edge without the need for internet access or cloud capabilities.