Only one out of four members in the team had deep knowledge of Artificial Intelligence (AI), but that did not stop them from developing an AI model that can accurately recognise and classify human gestures and poses in images.