Build on the presence layer.
We’re building InterfacesOS as the interface between humans and AI. Some of its primitives are ones we want in as many hands as possible: inside your devices, or inside what you’re building.
For device companies
You have the hardware, the supply chain, and the customers. What makes a device feel intelligent is the layer on top, and that’s the layer we build. Rather than becoming dumb glass, your devices can run one continuous intelligence that senses, remembers, and acts. We work with a small number of device partners who want their products to be AI-native, not AI-adjacent.
For developers
We’re opening the primitives behind InterfacesOS so you can build with them directly, including:
- Silent speech: talk to AI at the speed of voice, without making a sound.
- Agentic phone control: see the screen, operate apps, and run tasks in the background. Available today, open source, as OpenPhone.
- Memory & presence: the context layer that lets an agent stay useful across time and surfaces.
API access is early and invite-based today. If you have an integration in mind, tell us what you want to build.

