Your comments

When will you integrate this with client API calls so that we can programatically generate the speech?