My initial interview was via phone and was followed up with a test where I would use their system to listen to the a.i. audio and make corrections for pronunciations and pauses.