Bangladesh - 1

A Multilingual Context-Aware AI Moderation and Cyberbullying Shield powered by Meta Model API (muse-spark-1.3) and a multi-layered offline fallback engine. It accurately detects hate speech, threats, and subtle sarcasm (such as complex French and regional idioms) across 100+ global languages. It solves the critical limitation of traditional AI filters that fail to recognize sarcastic or hidden toxic phrases, providing a robust, real-time moderation solution through an interactive Gradio UI.