Yep, doable
You’d handle it by changing the voicemail content at runtime and tying it to call state (first call, second attempt like that). Each outbound call can generate a different TTS message using vars from the contact or previous attempts.
As long as you’re tracking attempts properly, it works fine.