Natural voice
Wake-word activation, local speech recognition and responsive conversational interaction.
A self-hosted personal AI robot combining natural voice interaction, local intelligence, expressive behaviour, computer vision and governed access to powerful AI tools.
TARZ is an independently developed robotics project exploring how conversational AI, local computing, sensing and physical interaction can work together as one coherent system.
Unlike a conventional smart speaker, TARZ is intended to understand spoken requests, interact naturally, perceive its environment, express personality through sound and visuals, and eventually interact physically with the world around it.
Core robot functions are designed to remain locally controlled, with cloud AI used selectively where it adds capability rather than becoming the robot's only operating dependency.
TARZ combines embedded robotics with modern conversational AI while keeping safety-critical behaviour under local control.
Wake-word activation, local speech recognition and responsive conversational interaction.
Raspberry Pi-based processing keeps essential speech, sensing and robot behaviour available locally.
Camera-based perception enables future awareness, tracking, recognition and environment-aware behaviours.
Displays, audio, lighting and character behaviours give TARZ a more physical and expressive presence.
Sensors, motors and embedded controllers provide the foundation for movement and real-world interaction.
External AI and software tools can be accessed through explicit controlled interfaces rather than unrestricted robot-level access.
TARZ separates conversational intelligence from the hardware safety systems responsible for controlling the physical robot.
TARZ is currently under active development. Voice, AI, interface, sensing and physical robot systems are being developed and tested incrementally before integration into the finished robot.