AI Receptionist for PracSuite Clinics

A caller rings a multi-discipline PracSuite clinic and says, "I've done something to my shoulder." That sentence could be a physiotherapist, an exercise physiologist, a remedial massage therapist, or, if there is numbness down the arm, something that should be looked at sooner. A receptionist who knows the clinic sorts it in a moment. A generic voice agent books the first physiotherapy slot it can find and gets it wrong often enough that the clinic stops trusting it. The whole job on a multi-discipline line is that first act of sorting, and it is the one most systems skip.

We build the PracSuite connection, and this is about that sorting: turning a described problem into the right practitioner, the right slot, and a booking written into PracSuite.

The difficulty is not booking; PracSuite books fine. It is deciding what to book from what the caller actually says, which is rarely the name of a service. 2care resolves the likely discipline and practitioner from the described problem, checks whether anything in the description should be seen sooner or by a clinician, and offers the right slot, all before it writes anything.

What the caller says	What 2care resolves	Where it books or routes
"Something's wrong with my shoulder"	A musculoskeletal problem	The right practitioner's next suitable slot
"My physio wants me doing exercise classes"	A referral into a class or program	The class booking, in the series
"Numbness down my arm since this morning"	A description that needs a clinician	Routed to a person, not booked blind
"I need to move next week's session"	An existing plan to reschedule	A new slot inside the plan window
"Do you take my health fund"	A cover question	Answered from the clinic's setup

Where the description genuinely could be two disciplines, the agent asks one more question rather than guessing, and where something in it should be seen by a clinician, it leaves the booking path and routes the call. Guessing the discipline wrong is not a small error on an allied health line: it wastes the patient's visit and the practitioner's hour, and it is the mistake a described-problem match is built to avoid.

What a good receptionist brings to this is not speed but knowledge of the clinic: which practitioner takes shoulders, who has a waitlist, which service a health fund actually rebates. That knowledge is exactly what the clinic has already encoded in PracSuite, and the agent reads it rather than improvising, so the sorting reflects how the clinic actually runs rather than a generic guess about what a shoulder needs. The difference between a booking that fits the clinic and one that merely fills a slot is that encoded knowledge, applied on every call at once.

How a booking reaches PracSuite, and how fast

Underneath the sorting is one mechanism. 2care connects to PracSuite through its API, authorised with a scoped credential set up during onboarding, with nothing installed at the clinic. Availability is read live per practitioner and appointment type, so an offered slot genuinely exists, and the booking is written against the resolved patient and validated by PracSuite against its own rules, so a slot taken during the call fails cleanly rather than being assumed free.

Speed is what makes the sorting invisible. A reply lands in roughly 480 milliseconds, p95 under 700, from a voice stack we run ourselves rather than one stitched together from outside services, so resolving a discipline and reading PracSuite live both fit inside a single conversational beat. Uptime sits at 99.9 per cent across 1000 or more simultaneous calls, and an idempotency key on the write means a dropped packet never doubles a session. The endpointing waits for a finished thought, not just a gap, so a caller with a long injury history is not clipped mid-sentence by an impatient timer.

When a described problem is not a booking at all

The reason the sorting has to be careful is that some of what arrives on an allied health line is not an allied health problem. Numbness down an arm, a sudden severe headache, chest tightness during exercise, a wound that looks infected: these are descriptions that should reach a clinician or a GP, not the next physiotherapy slot.

2care treats the boundary as the first question, not the last. Before it settles on a discipline, it checks the description against the clinic's escalation triggers, and anything that crosses one leaves the booking flow and routes to a person in seconds, with what the caller said attached. The agent does not assess or advise; it recognises that a description sits outside what it should book and hands it on, which is the same recognition-not-assessment line the whole platform is built on. Getting that boundary right is what makes the fast, confident booking on the other calls safe to trust.

Booking the plan, and catching the drop-off

Allied health income is the sum of hours booked, and most of those hours live in plans, not single visits. Where PracSuite supports it, 2care books a plan of care whole in one call, placing each session at the practitioner's cadence, so a patient leaves with the course in the diary rather than the first visit and an intention to ring back.

The other half of fill rate is the patient who was coming and then does not. Because a cancellation and a no-show are events the agent sees against the PracSuite diary, they become an action the same day rather than a line on a month-end report: a freed hour is offered down a priority waitlist within minutes, and a patient who stalled partway through a plan is called back while the gap is small enough to close. Catching that within a day rather than a month is where a surprising share of a clinic's lost revenue actually sits.

There is a discipline-specific twist here too. A patient who drops out of a physiotherapy plan and one who finishes a course of massage are not the same follow-up, and the outreach reflects that: the first is a rebooking to keep a plan on track, the second an invitation back when it is clinically sensible, both drawn from what PracSuite already holds about the patient rather than a blunt "we miss you" to everyone. Matching the follow-up to the discipline is what stops reactivation reading as spam and keeps it as a service.

What 2care changes for a multi-discipline clinic

A PracSuite clinic is not short of a practice management system. It is short of a phone that can sort several disciplines on the fly and fill the diary without another receptionist. 2care answers every call, resolves the discipline, books against PracSuite's own availability and rules, and writes the result back, in one native pipeline rather than a chain of tools passing a patient between them.

What a clinic feels is not a faster booking on the calls it already answered. It is the after-hours enquiries that used to reach voicemail now sorted and booked into the week, the plans booked whole instead of piecemeal, and the wrong-discipline bookings that used to waste an hour largely gone. 2care keeps a multi-discipline phone as orderly as the PracSuite diary behind it, at any hour and any volume.

The compounding effect is the one clinics notice a month in. A phone that answers every call, sorts each correctly, books plans whole and fills cancellations the same day does not just save time on the calls a receptionist was already taking; it recovers the hours that used to leak away unanswered, unsorted or unfilled, and those recovered hours are pure margin on a business whose cost base barely moves whether a slot sells or not.

The record, and the rules that apply

Before anything is written, the caller is matched to their existing PracSuite record and confirmed on a date of birth, so a returning patient keeps one file; a genuinely ambiguous match goes to a person with the candidates gathered, using patient matching, rather than a guess that splits a treatment history across two records. Intake is captured in the same breath and written to PracSuite as structured fields, not a note someone retypes, and the same capture feeds patient intake whether the patient rang, messaged or filled in a form.

Because a booking builds a transcript, matches a patient and writes to PracSuite, it processes health information rather than passing it through. Under HHS guidance for US work and the equivalent duties elsewhere, that makes the vendor a business associate or data processor rather than a conduit, so a signed agreement is in place from the first call, the data is held in region and encrypted in transit and at rest, the models behind the conversation never train on it, and every write is reversible. The compliance page sets out the detail.

Where this is right, and getting live

This is right for a clinic that runs more than one discipline off one number, keeps its practitioners, services and plan cadences current in PracSuite, and has a clinician to take the calls that should not be booked blind. Given those, the benefit lands from the first day the line goes live, and a busy Monday, when a single receptionist cannot both answer the phone and sort it, is exactly when it earns its place.

Getting there is three moves over weeks. The connection is authorised to read availability and write bookings through the PracSuite API. The clinic's setup is mapped: practitioners and disciplines, appointment types and durations, the cadence of a plan of care, and the route for a call that needs a clinician. Then the line is tested against a sandbox with the team listening before it is switched across, and the discipline-matching is tuned over the first days against real calls, each change backed by the trigger every escalation records rather than a hunch.

Frequently asked questions

How does it know which discipline a caller needs?

From the described problem, not a menu. It resolves the likely discipline and practitioner from what the caller says, asks one more question where two are genuinely plausible, and routes to a person where the description should be seen by a clinician rather than booked blind.

Does it book a whole plan of care at once?

Yes, where PracSuite supports it. Each session is placed at the practitioner's cadence with a reminder against each, rather than the first booked and the rest promised, which is what keeps a plan running to completion instead of thinning out after the first visit.

Can it fill a cancellation?

Yes. When a cancellation frees an hour, the agent offers it down a priority waitlist within minutes and writes the result into PracSuite, so a slot that opens in the morning is filled by the afternoon rather than lost to the month-end report.

Will it create duplicate records in PracSuite?

No. A returning caller is resolved to their existing file first and confirmed on a date of birth, and where two files are equally likely a person takes over rather than the agent guessing, so a treatment history stays on a single record.

Is patient data ever used to train models?

No. A patient's information is never used as training data for any model in the pipeline. It is encrypted throughout, held in region, kept only as long as the record requires, and covered by the agreement that governs every PracSuite call the agent takes.

Hear it sort a described problem

The call worth judging on a PracSuite line is the vague one: a caller who describes a problem rather than naming a service, sorted to the right discipline and booked, with anything that needs a clinician routed instead.

See how the same connection runs on other systems across integrations, or hear a described problem sorted and booked on a live call when you book a demo.