MILabs
Integrations

Best Practice AI integration: what AI can do with clinic software

How AI can help with scheduling, intake and admin around Best Practice, the clinic software many Australian GP clinics run.

Updated 2026-06-24

Best Practice: what AI can do with it

Best Practice is software that many general practice and allied health clinics in Australia use to run their day to day work. Clinics keep schedules, patient details and correspondence in a system of this kind, and much of the repeated admin sits around that record. AI helps most with that surrounding work. A clinic might use a language model to draft reminder text from appointment details, sort incoming requests into a queue for staff to review, or check a form for missing fields. Each use depends on what the clinic may share, which data a tool can read, and who approves the output. Start with a short list of tasks that repeat daily and have clear inputs and outputs. Then check whether a connection can be made with the software as installed, and whether staff will trust the result. Most of the gain comes from less copying between screens and fewer items waiting in a queue. This page covers what AI can do inside and around Best Practice, and where the limits sit.

Best Practice in plain language

Clinic management software holds the working memory of a clinic's day. Bookings, patient details, clinical notes, billing and correspondence all pass through one system that staff open dozens of times a session. Connecting AI to that kind of software means placing a language model or a small automated process beside it, so that some of the reading, sorting and drafting happens without someone retyping information from one place to another.

A model reads text, such as a referral letter or a free-text message, and returns a structured result, such as a suggested category, a summary or a draft reply. A workflow tool then decides what happens next. Whether that workflow can reach into the clinic software at all depends on what the vendor exposes, which can vary by edition and by how the installation was set up.

Two questions shape every project. The first is what the model is allowed to see, since patient information carries privacy obligations under Australian law and clinic policy. The second is what the output is allowed to do. A draft that a person reads before it is sent is a very different thing from a record that changes on its own. Most sensible designs keep a human in the step where the result touches care or money.

What Best Practice does

Best Practice is clinical software published by Best Practice Software. Its users include general practice and allied health clinics, along with some aged care and disability services that run similar systems for scheduling and records. Front desk staff, clinic managers and clinicians rely on it for bookings, records and day to day administration, so changes to how it is set up affect many routine tasks.

Where AI helps

  • Drafting patient reminder and recall messages from appointment details, for staff to check before sending.
  • Sorting incoming referrals and correspondence into a review queue so clinicians can see priorities first.
  • Checking intake forms for missing or inconsistent answers before they reach the front desk.
  • Summarising call notes and after-hours messages into a short handover for the next shift.
  • Answering routine questions about opening hours and forms through a chat or phone assistant, with escalation to staff.
  • Turning clinic reports into plain-language summaries for managers.

Easy wins around it

  • Access to clinical records depends on the permissions the vendor exposes and how the clinic has configured its installation.
  • AI output is not a clinical decision and should not be written into a record without review by a clinician.
  • Sending patient details to an external AI service must comply with Australian privacy law and the clinic's own policies.
  • Real-time write access to bookings or notes cannot be assumed for every installation.

How AI connects to it

  • Options for connecting to the product depend on the edition and setup the clinic runs, so confirm with Best Practice Software which data can be read and written before planning a build.

    Source: bpsoftware.net
  • Where the software can produce a report or file, that output can feed a review step, provided the clinic approves the data flow.

    Source: bpsoftware.net

Builds that fit

AI tools that fit

Best Practice in practice

Inside a clinic that uses this kind of software, the work usually begins with a queue. Messages arrive by phone, online form or patient portal, and each one needs a category before a person can act on it. An AI step can read the text, propose a category and a short summary, and place the item in the right queue. A receptionist or nurse then checks the suggestion and moves on. The gain lies in the sorting, while the decision stays with the staff member who owns that queue.

A second shape is the reminder cycle. Staff already know which patients are due for a recall, and the chore lies in writing and sending the messages. A workflow can gather the relevant appointment details, pass them to a model for a plain-language draft, and hold the result in a list for a nominated officer to approve. Wording is set once in a template, so tone stays consistent across the clinic.

A third shape is intake. Forms arrive with blank fields, answers that conflict with each other, or details filed under the wrong heading. A check step can flag these before the form reaches the front desk, so the person who rings the patient knows exactly what to ask. The flags prompt a person to act and decide nothing on their own.

good questions, straight answers

Frequently asked questions

Can Zapier, Make or n8n connect to Best Practice?
These tools connect to many business platforms, but whether one can reach Best Practice depends on what interface the installed product and edition expose. That information comes from Best Practice Software, not from a tool directory, so the clinic should ask the vendor. Once a route is confirmed, a workflow in Zapier, Make or n8n can move data between the software and other services, subject to the clinic's privacy and access rules.
What does AI not do well with Best Practice?
AI is weak at anything that needs clinical judgement, such as deciding whether a symptom is urgent or changing a plan. It can also be unreliable when a task depends on fields that are inconsistent, out of date, or spread across several screens. Language models can produce confident text that is wrong, so drafted messages and summaries need a named staff member to check them before they are sent or filed.
Which tasks in Best Practice suit automation?
Tasks that repeat on a fixed pattern suit automation best. Examples include reminder messages for upcoming appointments, checking that intake forms are complete, and routing a standard request to the right queue. Work that is rare, emotional, or clinically sensitive is a poor fit, and should stay with a person. A short audit of one week's tasks is usually enough to show which ones repeat often enough to be worth automating.
Can an AI assistant write to Best Practice?
Whether an AI assistant can write into Best Practice depends on the interfaces the installed product exposes, and that should be confirmed with Best Practice Software before any design. Where writing is possible, a sensible pattern is for the assistant to prepare an entry that a staff member reviews and saves. Direct writes to clinical records without review are a poor choice, given the risk of errors and the duty of care.
Who looks after the connection to Best Practice?
Ownership should be settled before anything is built. The clinic usually names an internal owner, such as a clinic manager or office lead, who approves access and changes. The software stays under the vendor's control, while the workflow built around it needs a named owner. Someone should review the connection on a regular schedule, and the clinic should keep a record of who has access and why.
let's make something happen

Workflow automation: how we can help

Choose how you want to work on this with MILabs. Both options start with an enquiry.

Discovery is free, no obligation from the first chat.

Let us build it for you

We build it and look after it.

Build it yourself, we'll help

You build it with our direction, teaching and a weekly check-in.