The operating layer for
appointment-based
service businesses.
We answer your calls, recognise the caller, load their history, check real availability with buffer and conflict rules, book them in, and run every confirmation, reminder, recovery and review that follows. One system governing every call, text and appointment in your business.
Sits above whatever booking platform you already use. Built to Canadian privacy standards (PIPEDA/PHIPA-aligned), and HIPAA-aligned for US deployments. Responds automatically in the caller's language across 30+ languages.
Live call flow · 00:00:02.431
~2s end-to-end · voice to booked
§01 · Every call
What happens between
“hello” and booked.
A single call triggers a sequence that most clinics stitch together with three staff members, two systems and a notepad. We run it as one governed process, and we log every step.
The sequence below is the actual execution order. Not marketing copy. If a step fails, it is written to the record and retried or escalated to staff with full context attached.
- +00:00Inbound
Call received
Inbound number matched to caller profile by E.164 lookup.
- +00:01Inbound
Language detected
Caller speaks French Canadian. Agent responds in French.
- +00:02Inbound
Memory loaded
Last visit 47 days ago · usual service: 60-min deep tissue · preferred RMT: Sofia.
- +00:18Booking
Availability checked
Real-time read across staff hours, service duration, buffer, Google Calendar conflicts.
- +00:22Booking
Slot offered
Thursday 2:30pm with Sofia · 60 min · $140.
- +00:36Booking
Booking written
Written to booking platform · duplicate check passed · booking_id generated.
- +00:37Confirm
Confirmation SMS
Sent in French · three-step delivery audit logged (planned → sent → delivered).
- +24hConfirm
24-hour reminder
SMS reminder with reschedule and cancel links.
- +2hConfirm
2-hour reminder
Final reminder with directions and parking note.
- postLifecycle
Staff check-in
Staff prompted to confirm attendance · response drives the next action.
- +1dLifecycle
Review request
Sent only if attended · tone calibrated to customer history.
- day 30Recovery
Cadence watch
Customer cadence is 28 days. They are 2 days overdue. Tier 1 nudge triggers.
- day 60Recovery
Recovery tier 2
Standard recovery message · personalised by last service.
- day 90Recovery
Final attempt
Tier 3 · after this, customer marked lost and removed from active pipeline.
Every step has a written record. Every message has a three-step delivery audit (planned → sent → confirmed). If anything fails, we log the failure and escalate with the full trace attached.
§02 · Voice agent
The voice agent that actually closes.
Purpose-built for appointment-based service businesses. It reads real availability, respects your buffers and staff rules, and treats returning callers like regulars, not strangers.
Multilingual by default
Detects the caller’s language from their first sentence and responds in that language. Over 30 languages supported, including English, French, Spanish, Mandarin, Cantonese, Portuguese, Arabic, Hindi, Punjabi, Tagalog and Italian. The customer never has to ask.
Real availability, not fuzzy
Reads live availability from the booking platform at the moment of the call. Accounts for service duration, staff hours, pre- and post-buffers, and Google Calendar conflicts. Never offers a slot that cannot be booked.
Memory on every call
Knows who is calling before they speak. Last visit, usual service, preferred practitioner, no-show history, file attachments already on record. Returning customers are offered their usual booking, not interrogated from scratch.
Live transfer to the owner
The agent recognises when a call should be a human conversation (complaints, package questions, sensitive cases) and transfers the call live, with a one-line context handoff so you never start cold.
Sample transcript · language auto-detected
Confirmation SMS sent in Spanish · 24h reminder scheduled
§03 · Lifecycle automation
From booking to dormancy, handled in one chain.
Confirmation, two timed reminders, post-visit staff check-in, no-show tiering, review request, cadence watch, three-tier recovery. Each step is triggered by a real event, not a calendar, and each message is delivery-audited end to end.
All of this runs on top of the booking platform you already have. If a customer replies, we route the reply back into the same conversation thread with full history intact.
- Booking confirmedMAIN
SMS in caller's language
- 24h reminderMAIN
Reschedule · cancel links
- 2h reminderMAIN
Directions · parking
- VisitMAIN
Staff check-in prompt
- No-show tier 1BRANCH
Polite rebook message
- No-show tier 2BRANCH
Softer tone
- No-show tier 3+BRANCH
Staff alerted, no auto-msg
- Review requestMAIN
If attended
- Cadence watchMAIN
Statistical · per customer
- Recovery tier 1RECOVERY
Personalised nudge
- Recovery tier 2RECOVERY
Day 30 standard
- Recovery tier 3RECOVERY
Day 60 final · then lost
Three-step delivery audit
Every outbound message is logged as planned, then sent, then confirmed. If any step fails we see it.
Replies stay threaded
Customer replies land on the same conversation, with full context. No fragmented inbox.
Nothing without a reason
No generic drip cadence. Messages fire on real events: booked, confirmed, attended, overdue.
§04 · Patient-driven profiles
Patients build their own file.
Insurance cards, referral letters, intake forms, progress photos. The patient texts them in as MMS. We recognise who they are by phone number, attach the file to their persistent profile, and make it visible on the dashboard and to the agent on every future call.
No portal logins. No email back-and-forth. The intake bottleneck disappears. No other system on the market does this.
Patient sends via MMS
Dashboard profile · live
Elena Ramos
+1 (***) 555-0143 · profile #4812
Patient-submitted files
4 on file
Blue-Cross-Gold.jpg
received 2 min ago
Patel-referral.pdf
received 1 min ago
Intake-2026-04.pdf
received last week
Pre-visit-photo.jpg
received last visit
On the next call, the agent will say
“I can see you've already sent in your insurance card and Dr. Patel's referral, so you're set for Thursday, nothing else needed at the front desk.”
Attached on the phone the patient uses · persistent across sessions
Attach to the right file automatically
We match by phone number. No logins, no patient IDs, no upload link the patient loses.
Visible where it matters
On the dashboard for the front desk. To the voice agent on the next call. To the SMS agent on the next message.
Agent-prompted when missing
If a document is expected before a visit, the SMS agent nudges the patient to send it in, with context.
§05 · Dashboard
One view of what the system is doing.
Bookings, customers, conversations, message delivery, automation pipelines, staff performance, recovery conversion. All of it in one dashboard, updated live as calls come in.
This is only a preview, click a tab below to switch views. Numbers shown are sample data and the live dashboard is fully interactive.
Overview
Bookings
142
128 completed
Revenue
$19,840
From completed
Calls
228
92% resolved
Messages
612
99.1% delivery
Booking activity · 30d
Lifecycle
- Active · 120
- New · 72
- Due · 51
- Lost · 24
What ABI did this period
- Answered 228 phone calls · 92% resolved without staff
- Booked 142 appointments · 18 recovered from dormancy
- Sent 612 messages · 99.1% delivery rate
- Flagged 7 no-shows to staff after tier-1 rebook attempts
The dashboard is built on the same data the agent uses, so what you see is what the agent sees on the next call. No second source of truth.
§06 · Competitor Watchdog
We watch the ten things your competitors do that actually matter.
Continuous monitoring of nearby operators across ten structured categories. Changes are detected by hashing page state, filtered for actionability, and delivered as short insights with a recommended response.
A supporting feature, not the headline. Runs quietly in the background once a client is live.
Read the full picture on Intelligence§07 · Integration
Above your booking platform, never in place of it.
We write into Square, Vagaro, Phorest, Jane, Google Calendar, or any platform with an API, using a thin adapter per platform. Your team keeps using the same booking software they use today. We govern the layer above it.
If your platform doesn't have an adapter yet, we build one as part of deployment. The rest of the stack stays unchanged.
Next step
Thirty minutes. We show you what it would do for your business.
We look at your calls, your tools, your hours, and your client flow. You leave with a concrete picture of where we would plug in, what the first week would look like, and whether it is worth it for you.
No commitment · free