Independent project / 2026
MRBD Voice AI
A lightweight voice-and-card experiment for Meta Ray-Ban Display, exploring what information deserves to appear in front of your eyes.
Why I am building it
Smart glasses should not inherit a phone interface at a smaller scale. The display is smaller, input is limited, and interruptions are more costly. MRBD Voice AI tests a lighter pattern on real hardware: show one useful card, complete an action by voice or a subtle gesture, then return attention to the world.
Current work
- Designing information density and D-pad navigation for a 600×600 glasses display.
- Exploring device pairing, voice input and short feedback loops.
- Running the service on an isolated subdomain and container.
- Documenting limitations around local ecosystem support, privacy and wearability.
Design principle
The interface should help someone reach for a phone less often, not create another screen that demands continuous attention. Each view gets one job, state remains scannable, and failures provide a clear recovery path.
Status
The project is in active development. The public URL supports pairing and staged testing rather than representing a finished product.