This should be fairly easy as it is simply a parameter in the API request and would allow us to select which voice to use.