Goal: To develop an intelligent system capable of converting hand gestures (American Sign Language) into audio output in real time, utilizing computation power of smart phones and unobtrusive (wear and forget) wearable technology. Description: The aim is to support people with partial or total speech impairment and elderly citizens with speech or motor impairment with a medium to interact with smart devices. The idea is to use gesture classification using Hidden Markov Models, LSTM or Bayes linear classifier for classifying gestures (Sign language gestures) and using Mary Text to Speech by DFKI to convert the classified gesture into speech.