Healthcare
From first contact to attended visit, with no manual coordination
Three intake channels, one pipeline, and a system that decides its own next move based on how the patient booked and whether they turned up.
Client: A clinic and aged-care provider
System flow and narrative
Booked via: Voice AI. Outcome: Attended. Path: Inbound call, then Routing and logic, then Unified calendar, then Patient record, then Confirmation call, then SMS confirmation, then Monitoring, then Day-before reminder, then Attendance check, then Satisfaction check-in, then Recall schedule.
From first contact to attended visit, with no manual coordination: system flow
Edges: live is a thicker, full-opacity, travelling dashed line; not taken is a thinner, dimmed, solid line. Nodes: live is a thicker, full-opacity outline; not taken is a thinner, dimmed outline. Color is never the only signal.
01
Three ways in, one pipeline
Patients reach the clinic through whichever channel they already prefer: a phone call handled by a voice agent, a web form, or a chat bot that pre-screens the request before booking. All three feed the same n8n pipeline rather than three separate workflows staff have to reconcile by hand. The single field that matters most is which channel the booking came through, because that one field is what every later branch in the system reads to decide its next move.
02
Booking and calendar sync
Whichever channel a patient books through, the appointment lands in one unified calendar, and n8n listens for the booking event regardless of source. The record carries patient details, service type, time slot, booking channel, and the assigned provider, so nothing downstream has to re-derive context that was already captured at intake. Double-booking and slot-conflict checks run automatically before the booking is confirmed, so a scheduling clash is caught before a patient ever hears back.
03
The confirmation call
Every booking, on any channel, triggers an outbound voice call that confirms the date, time, and provider and asks the patient to confirm or reschedule. The result, confirmed, no answer, or reschedule requested, is written back to the patient record automatically. If the call goes unanswered after two attempts, the system falls back to an SMS confirmation instead of leaving the booking unconfirmed, so a patient who cannot take a call is not simply dropped from the pipeline.
04
Branch one: how they booked
Once a booking is confirmed, the system enters a monitoring state that runs until the appointment date, and this is where the first branch happens. A patient who booked by voice has already had a live spoken interaction, so the system sends a standard day-before SMS reminder and stops there. A patient who booked by form or chat has not spoken to anyone yet, so the system schedules an additional pre-visit voice call 24 to 48 hours out. The extra touchpoint exists specifically to close the engagement gap between a spoken booking and a self-service one, not to add a step for its own sake.
05
Knowing whether they arrived
Attendance is recorded through whichever path fits the visit: front-desk or provider staff marking a patient attended or no-show directly, or an automatic check-in signal from a kiosk, an SMS check-in link, or an EHR sync. Whichever path fires, the same status field updates the patient record. n8n watches for that status change and uses it to trigger the second branch point in the system.
06
Branch two: attended or not
An attended visit triggers a satisfaction check-in and feeds a recall schedule, so a completed visit becomes the start of the next one rather than the end of the interaction. A no-show triggers a same-day SMS asking to rebook; if there is no reply within a set window, a voice call attempts to rebook directly; if that still goes unresolved, the system creates a task for front-desk staff instead of simply logging the miss. A no-show is also fed into risk flagging so a pattern of repeat no-shows is visible rather than treated as a series of unconnected misses. The point of the escalation chain is that an unresolved no-show becomes a task a person has to close, not a patient the system quietly lets go.
The branch logic, as a table.
| Condition | Then | Why |
|---|---|---|
| Booked via is Voice AI | Routing and logic | |
| Booked via is Web form | Routing and logic | |
| Booked via is Chat bot | Routing and logic | |
| slot conflict checked first | Unified calendar | |
| booking source recorded | Patient record | |
| fires on every booking, any channel | Confirmation call | |
| unreachable after 2 attempts | SMS confirmation | |
| Booked via is Voice AI | Day-before reminder | already had a spoken interaction |
| Booked via is Web form or Chat bot | Pre-visit call | no spoken interaction yet, 24 to 48 hrs before |
| Outcome is Attended | Satisfaction check-in | |
| Outcome is Attended | Recall schedule | 3, 6, 12 month recall |
| Outcome is No-show | Rebook prompt | same day |
| Outcome is No-show | No-show risk flag | |
| Outcome is No-show | Rebook call | no reply within the agreed window |
| Outcome is No-show | Front-desk task | still unresolved |
The stack.
| Layer | Tools |
|---|---|
| CRM and pipeline | GoHighLevel |
| Orchestration | n8n |
| Conversational AI | GoHighLevel AI Agent, OpenAI, Anthropic |
| Voice AI | Vapi |
| Calendar | GHL Calendar, Google Calendar, Cal.com |
| Messaging | SMS, WhatsApp, Twilio |
| Task management | ClickUp |
| Alerting | Slack |
| Data | GHL custom fields, n8n data store, EHR sync |
Compliance notes.
- HIPAA-aware handling throughout: minimal data in transit and access-controlled fields on every patient record.
- An audit trail of every automated action, so any confirmation, reminder, or rebook attempt can be traced after the fact.
- Call recordings and transcripts stored with restricted access and a defined retention policy, not kept indefinitely.
- EHR access through authenticated APIs only, with no PHI held in third-party automation tools beyond what the workflow operationally needs.
Before and after.
- No-show handling
- Before: One reminder channel, no automated reschedule pathAfter: Dual-channel reminders paired with an automated reschedule flow
- Front-desk manual follow-up load
- Before: Every booking and no-show handled manuallyAfter: Only unresolved edge cases reach staff
- Booking channel flexibility
- Before: Phone-only or form-onlyAfter: Form, chat, and voice unified into one pipeline
- Time to confirmation
- Before: Hours, via staff callbackAfter: Immediate, via automated voice confirmation
- Recall and retention follow-ups
- Before: InconsistentAfter: Systematic, triggered automatically post-visit
This is one delivered system, published in full. See how we build AI automation
Start a conversation
Tell us what you’re building.
Share the essentials and we’ll reply within 4 hours with a real next step, not an auto-responder.
What happens next
- We reply within 4 hours, from a real person, not an auto-responder.
- A short scoping call to understand the goal, constraints, and timeline.
- A fixed-scope discovery sprint: a working prototype and a written estimate.
- [email protected]
- Phone
- +1 504 732 9446
- Office
- Austin, TX, United States
- Hours
- Mon–Fri · Async + scheduled calls
