Currently only gathers labeled speech data from speech impaired individuals in a very simple and disabled-friendly way. The data will be used in future to train speech-to-text models to help speech impaired individuals in the task of speech recognition.