FEELBOT is an AI-powered emotional interface system designed specifically for children, combining advanced prediction with safe self-expression. It analyzes real-time emotional signals—facial expressions, emojis, color cues, and behavioral inputs—captured during interactions with a personalized avatar in a virtual space. Trained on over one million classroom-based emotional records, its proprietary algorithm achieves more than 90% prediction accuracy. The system delivers personalized content such as book recommendations, voice narration, and chatbot messages tailored to each child’s emotional state. Created by a former teacher, FEELBOT reflects deep educational insight and provides a safe, intuitive way for children to express, understand, and regulate their emotions. Beyond individual support, FEELBOT fosters empathy, social connection, and awareness of others’ feelings. Its effectiveness has been validated in schools and public programs, while data protection methods, Split Storage ensure anonymity and trust. FEELBOT represents a child-centered bridge between emotional care and AI innovation.