Tele-operated avatar robots serve as a physical extension of the operator, facilitating active and meaningful communication with local users. Recognizing that non-verbal cues are critical for shaping user experience, our system is engineered to deliver rich expressive behaviors. It integrates HMD-based eye tracking to map the operator’s dynamic gaze in real-time, a tele-haptics interface specialized for social touch, and an AI-driven voice modulation system to maintain a consistent character persona. By synchronizing these multimodal signals, the system significantly enhances the naturalness of the interaction, thereby increasing user immersion and engagement.