What is it ?
OM is a physical body for your personal intelligence, designed to live in your home or office. It is calm, screenless technology that takes functions off your devices and interacts with you in natural, multimodal ways. The name stems from the phonetic sound of "home", while evoking the meditative chant of "om", grounding intelligence in a peaceful presence.
Functionally, OM reimagines the smart home interface through three core interaction pillars. It features Ambient Response units, wooden sculptures that visualize the AI's voice through warm, pulsing light, effectively turning the room itself into a natural interface. The system employs a Follow-Me architecture, allowing the agent’s presence to fluidly transition between rooms alongside the user, maintaining a continuous session without being tethered to a single device. Crucially, interaction is governed by a Privacy-First minimalist wearable pendant; unlike standard "always-listening" speakers, OM only wakes upon a discreet button press, ensuring digital solitude remains the default state of the home.
The project is currently in the active prototyping phase, addressing a key question in modern interaction design: How can we architect a privacy-first, voice-controlled home operating system that feels natural and unobtrusive?
My role & work
My role involves the end-to-end conceptualization and engineering of OM's internal architecture, spanning hardware selection, system design, and conversational interfaces.
Conversational & Interaction Design: I am designing the "personality" and interaction flows of the agent, focusing on natural language understanding (NLU) and context-aware responses. This involves integrating OpenClaw to handle complex dialogue management, ensuring the agent can maintain state and context across multi-turn conversations.
System Architecture & Integration: I am architecting the central nervous system of the home, linking various subsystems (audio, media, automation) into a cohesive whole. This requires designing a robust event-driven architecture that allows the agent to autonomously manage inputs and outputs without a graphical interface.
Hardware Prototyping: I am leading the hardware design process, selecting and configuring edge computing components (Raspberry Pi 4B, Bluetooth peripherals) to create a stable, low-latency physical node. This includes solving challenges related to headless deployment, power management, and audio I/O routing to ensure a high-fidelity voice experience.
Iterative Development: As the project is in the prototyping phase, I am constantly iterating on the software stack and physical form factor, testing new integration strategies for "calm" technology that respects user attention.
All in all, this project serves as my playground for exploring the intersection of Embedded Systems, AI Interaction Design, and Smart Home Architecture.
