How we're automating 10,000 weekly conversations for a medical franchise

A network of 15 practices and 180,000 patients across one Auckland region. We are building the conversational assistant, chat and voice, that handles the reception load, escalates safely, and remembers every conversation.

A medical franchise in one region of Auckland runs 15 practices and clinics with around 180,000 patients between them. Across all of those front desks, the phones field the same relentless stream of calls: booking an appointment, ordering a repeat prescription, asking a quick general question.

At that size, it adds up fast, heading toward roughly 10,000 conversations a week. None of them are complex, but all of them tie up reception, and patients still struggle to get through.

Why do 15 practices and 180,000 patients overwhelm reception?

Because the volume is enormous and almost all of it is routine. A small number of front-desk staff per site cannot answer every call quickly, so phones stay busy, patients wait, and the team spends its day on bookings and repeat-script requests instead of the people in front of them.

Hiring enough reception staff to cover 10,000 conversations a week across 15 sites is neither affordable nor easy to scale.

What are we building?

One conversational assistant for the whole franchise, on two channels: a chat assistant on the website, and a voice receptionist on the phone. It handles appointment bookings, repeat prescriptions, general inquiries and first-line triage for every practice.

It connects to their Indici patient management system for appointments and patient records, and draws on a large Microsoft SharePoint knowledge base for everything a caller might ask about each practice.

How does it know when to step back?

This is the part that matters most in healthcare. The agent uses smart conversational logic to understand not just what a caller says but the tone and emotion behind it. When it physically cannot resolve a request, or senses the caller needs a person, it hands off to a human.

And if it picks up on an emergency, it escalates straight to 111. Routing the call to the right place, a person or an ambulance, is built into the design, not an afterthought.

Why save every conversation to a knowledge graph?

Most phone calls vanish the moment they end. Here, every conversation is saved into a knowledge graph, held behind a secure SharePoint location with clear permission logic over who can see what.

That means the context follows the patient through their care. During a consultation, for example, a doctor can pull up what the patient already discussed with the assistant, so they walk in with the history, not a blank page. One conversation, reused at every stage of the patient lifecycle.

What it is built to handle

Could this be your practice?

If your front desk is buried under routine calls, patients struggle to get through, or context gets lost between the phone and the consult, this is exactly the kind of problem we solve. We start with a Discovery Week: we find your single biggest win, scope it, and ship a first working piece fast, so you feel the value before any bigger decision.

Frequently asked questions

More case studies