Could use Nuance TTS. They support SSML to modify speech output.
Could use Nuance TTS. They support SSML to modify speech output.