Clarification

Between your text and the sequence, Aria asks the questions whose answer actually changes the plan. One at a time, with two to four clickable options, and always the option to answer freely.

What is asked, and what is not

The rule is explicit: never ask a question a professional has already settled. Nobody will ask whether you intend to anesthetize or take an impression. When hesitating between asking and interpreting, Aria interprets, and writes the assumption into the enriched plan so you can correct it.

Questions therefore land on real ambiguity: a vaguely named sextant, a bridge whose abutments are not adjacent, a tooth to extract or already missing, a provisional to decide on.

Your controls

• Undo my answer: takes back the last answer, without spending a call.
• Skip: forces Aria to conclude with what it already has.
• The mic: answer by voice.

Two visible guardrails

Aria cannot question you about a treatment you never mentioned: a question about endodontics, a graft or whitening absent from your text is blocked before it reaches you. And when dictation was expanded, an abbreviation turned into a procedure name for instance, the substitution is shown rather than silent.

The recap

At the end, one line per tooth with its treatments, a collapsible history of the questions, and the enriched plan text you can edit before confirming. That text, not your first sentence, is what generation works from.

Clarification can be turned off for the whole practice, and its rules adjusted per domain, in Settings under practice intelligence.