HomeOS

For AI to help with your physical home, it needs to understand the space, the things in it, and whether its proposed changes actually work. I built HomeOS: a shared model of your home that AI agents can understand, work with, and check their decisions against. You start with a room video. HomeOS combines a 3D reconstruction with measurements to create an editable scene. Agents have tools to inspect the room, find real products, build models, and test proposed changes. The important part is what happens when something fails. A placement blocks access. An object doesn’t fit. A model needs correction. Those findings guide the agent’s next attempt, while your original requirements stay fixed. You can see what changed, why it changed, and whether it passed. Our first demonstration focuses on furnishing and movement through a room. The broader vision is a persistent home model that agents can use to reason about lighting, accessibility, maintenance, and connected devices - with a way to test their decisions before acting. See how it works: https://homeos-judges-tiiny-host.tiiny.site/