Speech synthesis

Ohhh damn, today I devoted a day to this work, the work was carried out on accuracy and accuracy, because we are talking about the application interface, which hundreds of users currently use and the approach should be clear