They need to make the audio available to the user. Their text to speech engine never maages to get a decent transcript figured out. That or people are just yelling gibberish into the phone when I screen them. Until I can review the call audio I'll never know.