As a software engg myself i know its difficult to implement. Is there any standard pattern of stuttering? And how will and how long will the software understand the block. Is there any ROI for understanding such complex speech patterns? May be just 0.001% world population stutters. Unfortunately softwares are made to solve the problem of majority and not the micro minority.. 🙃