Bright Smile Dental AI Voice Agent
Clinics miss calls. During busy hours the phone rings while the front desk is with a patient, and every missed call could be a booking.
01 · Why it mattered
What the manual way was costing
- Missed calls are missed patients.
- Front desk gets pulled between the phone and the person in front of them.
02 · System concept
What I set out to build
A voice agent that picks up, understands what the caller wants, answers basic questions, handles the appointment part, and hands over anything it shouldn’t handle, with a summary so nobody starts from zero.
03 · Before / after
From manual steps to a system
Toggle between the manual process and the system. Click any step to see what it does, what tools it touches, and what happens if it goes wrong.
How it ran manually · depends on whoever is free
- 01Calls come in while front desk is busyby hand
- 02Missed calls go to voicemail or nowhereby hand
- 03Callbacks happen later, if at allby hand
- 04Details from calls are scribbled or forgottenby hand
04 · Decision log
Why I built it this way
D-01Knowing when to hand over
For a clinic, the agent being wrong is worse than the agent transferring. I designed the escalation and summary as part of the core flow.
D-02Grounded in real data
Answers come from patient and treatment data via Supabase / MCP, not from whatever the model feels like saying.
05 · Reliability layer
What can go wrong, and who steps in
- Where it happens
- Step 02 Understand intent
- What goes wrong
- Agent can’t tell what the caller needs
- What the system does
- Transfers / escalates instead of bluffing through it.
- Who handles it
- A person: front desk
Human-in-the-loop points
- Human in the loopBook or transfer
- Human in the loopAgent can’t tell what the caller needs → front desk
- Human in the loopCaller asks something clinical or urgent → clinic
- Human in the loopPatient or treatment data not found → front desk
06 · Evidence
What exists today
- Demo build of the voice agent.
07 · What this demonstrates
What this demonstrates
Designing a voice AI system with clear limits, data grounding and a human escalation path.
08 · Stack / integrations
What it’s built with
- ElevenLabsVoice agent
- SupabasePatient and treatment data
- MCPTool access for the agent
- → front deskAgent can’t tell what the caller needs
- → clinicCaller asks something clinical or urgent
- → front deskPatient or treatment data not found
More systems
Keep reading
RecruitTech ATS
A deployed applicant tracking app where every application becomes a candidate record moving through screening and interview stages.
- Application
- Candidate record
- Screening / status
- Interview stage
- Recruiter action
- +1 more
Founder Lead Intelligence & Outreach Automation
Finds businesses, extracts and validates contacts, qualifies them and drafts personalised outreach into a CRM, with a human review before anything goes out.
- Business source
- Extraction
- Email validation
- Qualification
- CRM / database
- +2 more
Similar workflow?
Got a process that looks something like this?
Tell me how it runs today and where it hurts. If it’s a good fit for automation I’ll tell you how I’d approach it. If it isn’t, I’ll tell you that too.
Tell me what you’re trying to automateOr reach me directly