Then it was definitely bots XD
I completely missed the obvious detail that you need the times voiced. I've been using uap (Unity Accessibility Plugin) fairly successfully. I don't even bother with the GUI aspect of it. I just use their Speak function to say out loud text strings. It works flawlessly for desktop builds. For WebGL builds it uses a voice that's not what I have setup in my TTS. Not sure what exactly I'm missing there. But it does say the lines anyways.