The caller reaches the product. The agent knows the account, can update a record the user confirms, and a person on your team can take over with the transcript still in front of them.
The call is tied to that tenant. Recording and consent depend on where the caller is. Your counsel should look at that before go-live. This note is not legal advice.
The build is voice as a feature of your SaaS, part of AI features for SaaS.
Yes. Any one of these can be the first feature, once the product and the tenant boundary are clear.
