GeniePod is open source home AI built to live where it belongs: in your home, not someone else's cloud.
OpenClaw proved something important: people do not just want access to AI. They want AI that feels personal, remembers context, and fits into daily life.
That demand is real. But for home automation, the trust model has to be different. A home AI does not just answer questions. It hears family conversations, remembers household context, learns routines, and can control physical devices. That makes privacy and safety architectural requirements, not product features.
GeniePod is built around that idea. It is not a generic chatbot in a box. It is a local physical AI ecosystem for the home: private by default, offline-first, subscription-free, and designed for safe control of real devices. GeniePod has four core layers:
- GenieOS: the custom Jetson-based operating system and hardware layer.
- Genie AI Runtime: local inference optimized for NVIDIA Jetson hardware.
- Genie Home Runtime: AI-native home automation, device control, and deterministic safety checks.
- GenieClaw: the agent layer for voice, memory, skills, routines, and household interaction.
If you have imagined using OpenClaw to automate your home, GeniePod is built for that exact direction, but with the missing trust model: local memory, local voice, local control, and safety gates before physical actions.
OpenClaw 🦞 proved the demand. GeniePod brings that idea home.