Their software tools allow for easy programming of mobile devices and apps to incorporate gesture recognition features. Their Gesture Recognition Engine algorithm captures motion data and trains a neural network to compare random movements to intentional movements, allowing for human behavior analysis, rep counting, technical and postural evaluation in training, and device control via Bluetooth. Their super-low-consumption high-accuracy engine can run offline on wearables, devices, and apps, or online on the cloud in a Gesture as a Service model.