Challenge
Most AI wearables prioritize productivity—recording conversations, extracting tasks, and pushing notifications—while ignoring how the user actually feels. Cloud-dependent processing introduces latency, privacy risks, and emotional misinterpretation. The challenge was to design a fully local, low-latency wearable AI capable of detecting emotional states in real time, maintaining contextual memory, and responding with empathy—without screens, cloud APIs, or constant user input.
Goal
Create a wearable AI that: Understands how something is said, not just what is said perates entirely on-device with zero cloud reliance Builds long-term emotional context through memory Responds briefly, calmly, and without interruption Preserves user trust through radical privacy
Result
Rework
