Description
The most damaging form of bullying, verbal bullying, has never been detectable at scale with human effort alone. Bullying Decoder changes that through AI-driven assistance. It reimagines existing classroom CCTV systems by activating unused audio feeds and transforming that data into a behavioural intelligence tool that captures what schools have always missed. Trained on hours of real classroom audio, the AI interprets tone, repetition, and emotional intensity to identify subtle patterns of harm across multilingual environments. It converts classroom conversations into anonymised, aggregated insights visualised on a dashboard, enabling counsellors to take grade-specific action rather than relying on generic sessions that lack impact. By transforming what was once dismissed as background noise into actionable insights, without ever identifying individual students, Bullying Decoder showcases a privacy-first use of AI, shifting classrooms from passive monitoring to active understanding.
