Can an AI call my clients in my own voice?
Yes. A coach records their own voice once, and ShoUp's AI agent uses it to place real phone calls to clients, on a schedule the coach sets, running the exact questions the coach wrote. The client hears their coach checking in, the way that coach always said they would.
It isn't a recording and it isn't a script read flatly. The AI agent listens to what the client says and responds to it, then logs the answer for the coach to read before the next session. If the call goes unanswered, it tries again. Nothing about the call tells the client "this is ShoUp", because to the client, it isn't.
How does the voice actually get cloned?
A coach records a short, clean voice sample once, in a quiet room on a decent microphone. That sample trains a voice model the AI agent uses on every call after that. It's set up once and it stays the coach's own voice from then on, not a generic AI voice with the coach's name attached to it.
Why does the voice have to be the coach's at all?
Because the relationship is the reason clients follow instructions in the first place. About a third of clients stick to a plan once they leave a coach's care, and in some settings it's closer to half who don't follow through at all.[2] A generic AI voice reading the same questions doesn't carry that relationship. The coach's own voice does.
Does the client know it's an AI?
If they ask, the coach tells them the truth, that's the honest answer, and the one that actually builds trust. What the client experiences by default is their coach's program calling on schedule, asking the questions their coach wrote, in their coach's voice. Nothing about the call announces a third-party platform, because there isn't one from the client's side. This is the white-label design ShoUp is built around.
Why does the call matter more than a text or an app?
Text and app-based check-ins wait for the client to open them. A call comes to the client instead, and it's a channel most people still pick up even when they're avoiding everything else. Reporting an answer to someone, out loud, on the day it's asked for, is the same mechanism behind why checking in works at all.[1]
What happens to the answers?
Every call gets logged: who was reached, what they said, who's on track and who's slipping. The coach reads a plain-English summary before their next real session, and the AI agent flags anything that needs a real conversation, a breakthrough or a client starting to fall off, so the coach's time goes to the moment that actually needs them.