The platform

A conversation engine and a scheduling engine, working as one.

CliniCall sits between the systems that hold your appointments and the patients who hold the other half of the information. It closes the loop that reminders leave open.

Where it sits

Middleware, not another system for staff to log into.

CliniCall is designed to work with the appointment book you already have. Nothing about the way your service books appointments needs to change for it to start working.

Your systems

  • Patient administration and scheduling
  • Clinic and clinician availability
  • Waiting lists
  • Contact details and consent flags

CliniCall

  • Conversational calling engine: two-way natural language, inbound and outbound
  • Backfill engine: real-time matching of freed slots to waiting patients
  • Rules layer: your clinical and operational policy, enforced
  • Staff console: live view, override and handover
  • Audit trail: every contact, decision and outcome logged

Your patients

  • Inbound calls to the practice
  • Outbound voice call
  • SMS
  • WhatsApp
  • Existing letters and channels, unchanged
  • A person, whenever it matters
Integration is an option, not a precondition. CliniCall runs perfectly well without a direct connection to your scheduling system, and at the lightest setting it needs no connection at all: we work from a list you provide, and hand back each agreed outcome for your team to enter. That is a few seconds at a keyboard in place of a call nobody had time to make, and it can be running while the integration conversation is still being had. Services that start this way add the connection later, once it has proved itself and the work can be justified. Nothing about the product changes when they do.
Engine one

The conversation is the product.

Automated calling has a bad reputation in healthcare, and it has earned it. Keypad menus, rigid scripts, and a dead end whenever the patient says something unexpected. CliniCall was built to be the opposite of that.

Natural language, both ways

Patients speak or type normally. “I’m away that week”, “can you do after five”, “my mother needs to come with me”, all handled in the conversation, not deflected to a phone number.

Channel escalation

Starts on the channel most likely to reach that patient and escalates if there is no response: SMS to voice to WhatsApp, on a schedule your service sets.

Human handover

Distress, confusion, a clinical question, or anything outside the rules ends the automation and passes to a person, with the full context of the conversation attached.

Multilingual by design

The conversation layer is not tied to English. For services with significant non-English-speaking populations, this is a reachability gain, not a nice-to-have.

Calling windows and volume limits

UK calling hours, per-patient contact limits and a configurable cap on simultaneous calls. The system will not surprise a service with a wall of outbound calls.

No fabricated certainty

Outcomes come from deterministic events, not from a model’s impression of how the call went. An unclear outcome is recorded as unclear and routed to a person.

Engine two

A freed slot is only recovered if someone else uses it.

This is the half most tools skip. Getting a patient to cancel early is useful; getting that slot filled before the clinic runs is where the capacity actually comes back.

The moment a slot is released, by a decline, a reschedule or a cancellation from any source, the backfill engine looks for the right patient to offer it to, and starts a conversation with them.

  • Your rules decide. Clinical urgency, waiting time, appointment type, clinician, site, travel distance and vulnerability flags are all inputs your team controls.
  • Offer, not assignment. The slot is offered and confirmed conversationally. Patients are never moved without agreement.
  • Race-safe. Several patients can be offered a slot at once; the first confirmation takes it and the others are told immediately and kept in place for the next one.
  • Written back. The appointment book is updated in real time, so the clinic list a receptionist sees is the true one.
The measure that matters: time-to-refill.

A slot released three days out is worth far more than the same slot released on the morning. CliniCall reports decline lead-time and time-to-refill from the first week, because those two numbers, more than any headline percentage, tell you whether capacity is genuinely being recovered.

We agree the target with you, measure against it, and show you the working.

In practice

Four products, and what each one actually does.

These are the same four you will see if we demonstrate the platform, under the same names.

Confirm & Backfill · Days before · the core of the platform

Re-confirm attendance, and refill what frees up

Circumstances change between booking and clinic. A one-way reminder cannot respond to that; a conversation can.

  • Ahead of the appointment, CliniCall checks the patient can still attend, escalating across SMS, voice and WhatsApp until it reaches them.
  • If they cannot, it reschedules them inside the same conversation, offering only what your clinical allocation rules permit for that patient: no callback, no queue, no second attempt from reception.
  • The freed slot goes to the backfill engine, which offers it to the next suitable patient under the clinical rules your team sets.
  • Urgency, vulnerability, waiting time and appointment type decide who is offered the slot. Your rules, not ours.
The difference: A cancellation stops being a lost slot and becomes a slot that moved to someone else who was waiting.
Configured to your clinical settings. How far ahead attendance is checked, how many attempts on which channels, which appointment types may be moved and to what, who is eligible for a freed slot, and the order urgency, vulnerability, waiting time and appointment type put them in.All of it is configuration your service owns and can change, not code we rewrite. Nothing runs until your team has set it.
For your team

Staff stay in charge, and can see everything.

CliniCall is only useful if the people running the clinic trust it. That means visibility by default and control at every step.

Live queue

What is queued, calling, answered, completed, failed or blocked, in one view.

Safe mode

New deployments start with calls requiring manual approval, until your team is ready to let it run.

Override anything

Pause a campaign, exclude a patient, stop a clinic, or take a conversation over. Every override is logged.

Full audit trail

Every contact attempt, patient response, decision and write-back, with timestamps, exportable.

Integration

Three ways to connect, and the lightest one starts now.

CliniCall does the expensive half of a rebooking whichever you choose: reaching the patient, agreeing a time, and confirming it. All integration changes is who enters the result, and how quickly the freed slot is back in play. Start at whatever level your IT position allows and move up when it suits you.

Level one, no integration
We work from a list you provide and hand each agreed outcome back for your team to enter. Nothing connects to your systems, so nothing waits on your IT department, and you can be live in days. Every conversation, decision and outcome is still logged and exportable.
Level two, secure file exchange
A scheduled export in and a controlled write-back path out. No supplier API needed, and the keying-in disappears. This is where most pilots run while a direct integration is scoped.
Level three, direct integration
Real-time read and write against the system that holds the appointment, so a slot released at 09:02 is offered at 09:03. This is where the return multiplies, and it is the destination rather than the entry requirement.
Scheduling systems
Designed to integrate with NHS scheduling systems, including TrakCare and SCI Store in Scotland. We are in discussion with established NHS systems providers about direct integration. Tell us what you run and we will tell you precisely how we would connect to it.
Interfaces we work through
Where a system exposes a standards-based interface we use it: HL7 v2 messaging, FHIR resources for appointments and patients, or the supplier’s own API. Where none of those exist, a controlled file exchange does the job until one does.
Integration model
Provider-agnostic core with adapters, so a new scheduling system is a new adapter rather than a rewrite. Contracts are published as OpenAPI, and your team can read them before anything is connected.
No silent writes
Reads can be automated. Writes are explicit, auditable, and switched on deliberately, never as a default.
Idempotent by design
Re-fetches, retries, webhook replays and restarts cannot produce duplicate calls or duplicate outcomes.
Hosting
UK-hosted, and deployable within your own infrastructure where your information governance requires it.
Built as middleware, deliberately. CliniCall’s architecture spans the full appointment stack, covering GP, outpatient, diagnostic and radiology, because a tool that only works in one system is a tool a service outgrows. Tell us which systems you run and we will tell you exactly how we connect to them.

See it against your own pathways.

Twenty minutes, your appointment types, your rules. We will show you where CliniCall would sit and what it would need from you.