Axon – AI that detects your behavioral patterns with evidence thresholds
Axon is an AI system designed to detect human behavioral patterns by applying evidence thresholds, aiming to identify and analyze recurring behaviors based on collected data.
Background
HGen is a Polish deep-tech startup that has developed software for analyzing human behavior and mental states. The core of Axon is a "behavioral foundation model" — an AI system that, instead of analyzing text content (what you say), focuses on how you speak, move, and gesture. It processes video, audio, and text to detect patterns linked to cognitive states (e.g., confusion, focus, fatigue, rapport). A distinctive feature is its use of "evidence thresholds" and explainability: the system shows which specific behavioral cues (e.g., eye movement, pause length, facial expression) triggered a conclusion, and how certain it is. This addresses a common criticism of "black-box" AI that cannot explain its reasoning. Axon is not a consumer product; it's a platform for enterprise or research use, aiming to power applications in recruitment, healthcare (e.g., early dementia screening), education, and human-machine interaction. The company positions it as a scientifically grounded alternative to hype-driven emotion AI, with published validation studies.