Upgraded fast-track AI speech model — improved voice naturalness and emotional expressiveness at high speed for scalable, responsive voice-enabled application development