A Voice V-Rep is not a black box with a voice. It is nine tabs, and each one is a decision you can see and change. Walking the anatomy is the fastest way to understand why configuration beats code: there is nothing hidden to reverse-engineer.

The nine tabs

Why nine, and not a config file

Every one of these is a decision a voice line has to make whether you expose it or not. Developer APIs make the same nine decisions; they just bury them in code and four vendor dashboards. The tabs are not a simplification of the real thing. They are the real thing, made legible to the person who owns it. See the full console.

Email is ten, chat is six

The channel decides the tab count. Email adds Email Connection, Monitoring, Tracking, Escalation, Knowledge Base, and Logs because an inbox has obligations a phone call does not, like an audit trail on every thread. Chat is the lean six because an embedded widget needs less. The core, LLM, Tools, Analytics, Memory, is shared across all three, which is why one workspace can run a voice line, an inbox, and a widget without three integrations.

The point of seeing it

An operator who can name all nine tabs can change any of them. That is the difference between owning a line and renting one you cannot touch. This is the case for configuration over code, made concrete.