In development

Meet MIYO

A small desktop conversational AI companion being designed as an ambient, screenless alternative to a smart speaker.

MIYO is in early development. Engineering begins once our first internship cohort is onboarded. Follow along or join us — see Careers.

MIYO concept render

Design principles

These guide the intended MIYO experience; they are not shipped features.

Privacy First
No Camera
Far-Field Mic
Voice Activated

The intended architecture

MIYO is planned as a proactive companion that can observe routines and initiate conversation, rather than only responding to commands.

Target hardware

A planned ESP32-S3 microcontroller with an OLED expression display and an I²S audio input/output loop.

Intended voice path

Wake-word detection and audio capture are planned on-device, with voice orchestration, reasoning, and task execution in the cloud.

Memory

The design includes persistent per-user memory so the companion can recall past interactions and preferences.

Design target

MIYO is being designed for a voice round-trip targeting sub-300 ms end to end.

Follow MIYO's development

For development updates, email us with “MIYO updates” in the subject line. This is a development-updates list, not a purchase, pre-order, or shipping waitlist.

Email for development updates
No camera is part of the intended design.