github·Robotics and devices Featured Pick
Android On-Device Jev Runtime
Quantized INT8 decision model running natively on Snapdragon NPU hardware inside an Android service without internet connection.
Open GitHub ⚡ 14ms on Snapdragon 8 Gen 3
Classify user screen state for accessibility voice navigation.
Jev sits between the incoming context and the next system action. Rather than generating lengthy, slow natural language that requires brittle regex parsing, Jev returns non-autoregressive, calibrated probabilities that downstream code can immediately execute.
01screen_intentPrimary action
02focused_componentConfidence score
03spoken_labelFallback route
Related Robotics and devices Builds
xpost
InspectVoice control over Fusion 360
I made this for fun. Wired a mic over Fusion 360: click, speak, Fusion runs the feature. @typesafeai ‘s Jev only decides if the utterance is a command.