Access restricted to Orthodontic Partners employees
Each new patient call is reviewed by AI against Orthodontic Partners' 10-category rubric. Each category is scored 1–5 (50 points max, displayed as 0–100). The AI reads the full transcript and scores each dimension independently, then generates specific coaching opportunities and supporting evidence quotes. Calls sourced from GoTo → Snowflake pipeline (Apr 8 – Jun 8, 2026). 309 new patient calls identified from 2,077 total inbound calls.
Risk reflects how likely the patient is to follow through and actually show up based on signals in the call — unresolved concerns, no appointment confirmed, hesitation, or an abrupt ending.
Specific moments in the call where a different response — a direct ask, addressing a concern, clarifying insurance — would have meaningfully improved the outcome. Each opportunity is assigned a priority based on how much impact it likely had on whether the patient booked.
Click a row to see that person's coaching breakdown and individual calls.
Click any row to open the full rubric breakdown, coaching notes, and call evidence.
| Date | Staff | Score | Outcome | Risk | Duration | Top Coaching |
|---|
Each call is scored across three dimensions. Scores reflect only what is observable in the transcript — the AI does not reward behaviors that aren't present, and does not penalize staff for caller-driven gaps. Composite = (Foundation × 0.4) + (Performance × 0.5) + (Hospitality × 1.0). If Foundation < 60, Composite is capped at 65.
Did the coordinator follow the required steps on every call?
Signals: Greeting includes practice name + coordinator name · Caller name captured · Reason for call identified · Basic info collected (DOB, address, contact) · Appointment offered · Appointment confirmed
How well did the coordinator run the call?
Signals: Follow-up questions asked · Agent leads conversation · Proactive scheduling (specific times offered) · Structured scheduling approach · Objection handling · Value connection (explains why to come in)
Did the call feel human and personal?
Signals: Empathy present · Caller name reused after capturing it · Personalization (references specific details) · Warm tone · Non-transactional language