Agent Case Study · Dental Group
Every call answered,
even the ones at 9pm
A dental group with four locations was losing patients it never met: calls rang out during lunch rushes, after 5pm, and every Saturday. We deployed an AI voice agent on their existing numbers — it answers instantly, books into their practice calendar, and hands anything clinical straight to staff.
Industry
Dental (multi-location group)
Company size
4 locations, ~35 staff
Region
Southeast US
Time to live
4 weeks, after-hours first
Client details anonymized under NDA — industry, size band, and timeline are accurate; identifying details are not shared.
Where they started
The challenge
Front-desk teams were answering the same five questions — hours, insurance, directions, availability, rescheduling — between checking patients in. Call answer rates during peak hours dropped below 70%, and after-hours callers got voicemail, which new patients almost never leave.
The group had tried an answering service; patients disliked reading from a script as much as staff disliked retyping the messages. Whatever replaced it had to book real appointments into their practice management calendar, not take notes about them.
The build
What we built
01Call-flow mapping
We listened to a week of recorded calls across locations, defined the routine intents worth automating, and set hard escalation rules: anything clinical, in pain, or upset goes to a human immediately.
02A voice patients talk to
Built on Twilio with a natural voice, the agent identifies itself honestly, keeps answers short, and transfers on request — no phone-tree mazes.
03Real booking, not messages
Calendar integration lets the agent offer actual open slots per location and provider, then confirm by SMS before the call ends.
04After-hours first, then overflow
It took nights and weekends for two weeks while staff reviewed transcripts daily, then picked up daytime overflow once trust was earned.
Ninety days later
The results
The unexpected win was the transcripts: for the first time the group could see exactly why people call, which reshaped their website FAQ and cut call volume further. Resolution landed at 71% — inside the 60–80% range we publish — with everything else reaching a human who now starts the conversation with context instead of 'can I put you on hold?'
Want this outcome in your business?
Start with a free AI audit: we map your workflows and show you where this agent pays for itself — before you commit to anything.