Ai agent how its work

Short summary

Collins Street Respiratory and Sleep Clinic is a fictional specialist medical practice in Melbourne, created as a demonstration case study for medical practices.

Like many healthcare providers, the clinic needs a safe and professional way to respond to callers when reception staff are unavailable. Patients may call about appointments, referrals, fees, results, prescriptions, location details, or general practice information. Some calls are simple administrative enquiries, while others must be handled carefully and passed to the clinic team.

FibreMax created an AI After Hours Assistant that answers calls outside clinic hours, gives a clear emergency notice, answers basic administrative questions from the approved clinic knowledge base, and takes structured messages for the clinic team.

The result is a safer and more consistent after hours call experience, while making sure the AI does not provide medical advice, interpret symptoms, confirm appointment availability, or make clinical decisions.

The situation

Medical practices receive a wide range of calls every day.

Some callers need simple information, such as opening hours, location, referral requirements, or what to bring to an appointment. Others need help with appointments, test results, prescriptions, clinical questions, fees, or patient specific matters.

For a busy medical practice, this creates a challenge outside normal reception hours. Callers still want to be acknowledged, but the clinic must also make sure information is handled safely and appropriately.

Collins Street Respiratory and Sleep Clinic is open Monday to Friday from 8.30 AM to 5 PM. It is closed on Saturdays, Sundays, and Victorian public holidays.

Outside those hours, the clinic needed a professional way to answer calls without creating the impression that the caller was speaking with a nurse, doctor, or clinical triage service.

The challenge

The clinic needed an after hours solution that could help callers without creating medical or compliance risk.

A standard voicemail can capture a message, but callers often leave incomplete details. Staff may then need to call back just to clarify the reason for the call, whether the caller is a new or existing patient, whether they have a referral, and what action they need.

At the same time, a medical AI assistant must be carefully controlled.

• It must not diagnose.
• It must not triage.
• It must not provide treatment advice.
• It must not interpret test results, sleep studies, lung function tests, scans, referrals, or specialist letters.
• It must not confirm appointment availability.
• It must not quote exact fees unless approved information is available.
• It must not make promises on behalf of the clinic team.

The goal was to create an assistant that could be helpful, but still stay inside safe administrative limits.

The goal

The goal was to create an AI After Hours Assistant that could.

• Give an emergency notice at the start of every call.
• Answer simple administrative and general practice questions.
• Explain the clinic hours, location, and contact details.
• Explain general referral requirements.
• Explain what patients should bring to appointments.
• Explain that fees apply and Medicare rebates may apply with a valid referral.
• Take a structured message when the caller needs follow up.
• Quickly move to message taking for patient specific matters.
• Avoid medical advice completely.
• Keep responses short, calm, and professional.
• Send the message to the clinic team for review and follow up.

The FibreMax solution

FibreMax created a controlled AI After Hours Assistant for Collins Street Respiratory and Sleep Clinic.

The assistant is designed to act like a polite medical reception assistant, not a clinical service.

At the start of every call, the assistant gives a clear emergency notice.

From there, the assistant can answer basic administrative questions using only the approved clinic knowledge base.

For example, it can answer questions about.

• The clinic name.
• The clinic location.
• The clinic phone number.
• The clinic email address.
• Opening hours.
• The main doctor.
• General areas of care.
• Referral requirements.
• What patients should bring.
• Basic fee and Medicare rebate information.
• The message taking process.

For anything patient specific, appointment specific, billing specific, or clinical, the assistant moves to message taking.

How the call flow works

The caller rings the clinic after hours.

The AI After Hours Assistant answers the call.

The assistant immediately gives the emergency notice.

The caller explains what they need.

If the caller asks a simple administrative question, the assistant answers briefly from the approved knowledge base.

If the caller asks about symptoms, results, medication, prescriptions, referrals, appointment changes, fees, or anything patient specific, the assistant does not attempt to resolve it.

Instead, the assistant offers to take a message.

The assistant collects one detail at a time.

The standard message details include.

• Caller name.
• Best callback number.
• Whether the caller is a new or existing patient, if relevant.
• The reason for the call.
• Preferred contact method, if offered.

For new patients, the assistant may ask whether they have a referral.

For existing patients, the assistant may collect the patient name, date of birth, and reason for contact.

For referrers or medical practices, the assistant may collect the practice name, provider name, patient name, referral related message, and return contact details.

Before sending the message, the assistant confirms the details with the caller.

Once confirmed, the message is sent to the clinic team.

The customer experience

For the caller, the experience is more helpful than voicemail.

• They are greeted professionally.
• They hear the emergency instruction immediately.
• They can ask simple questions about the clinic.
• They are not encouraged to discuss medical concerns with an AI assistant.
• They are guided into leaving a clear message for the clinic team.
• They do not need to guess what information to provide.

For the practice, this creates a more consistent after hours experience.

Instead of receiving vague voicemails, the clinic receives structured messages with the details needed for follow up.

The business outcome

With the AI After Hours Assistant in place, Collins Street Respiratory and Sleep Clinic has a safer and more consistent way to manage after hours calls.

The expected benefits include.

• Fewer incomplete voicemail messages.
• Better structured patient enquiries.
• Clearer follow up information for reception staff.
• A professional response for after hours callers.
• Reduced risk of callers receiving inappropriate medical advice.
• Consistent emergency messaging at the start of every call.
• Better separation between administrative information and clinical matters.

The assistant does not replace reception staff, nurses, or doctors. It simply helps the practice capture enquiries more clearly when the team is unavailable.

Before and after

Recorded voice sample calls

This case study can include one or more sample call recordings.

Suggested sample call topics.

• A new patient asking whether they need a referral.
• An existing patient asking about appointment changes.
• A caller asking about clinic hours, location, and parking.
• A patient asking about test results, with the assistant safely refusing to interpret results and offering to take a message.
• A patient asking about medication or prescriptions, with the assistant safely moving to message taking.

The recordings should show that the AI assistant is helpful but controlled. The main point is to demonstrate safe message taking, short responses, and clear boundaries around medical advice.

Why this matters for medical practices

Medical practices need to be accessible, but they also need to be careful.

Patients often call outside business hours with questions that feel important to them. A normal voicemail may not capture enough detail. A poorly controlled AI assistant could create risk by saying too much.

A FibreMax AI After Hours Assistant gives medical practices a practical middle ground.

• It can answer approved administrative questions.
• It can provide consistent emergency messaging.
• It can collect structured messages.
• It can support reception staff by improving the quality of after hours enquiries.

Most importantly, it can be designed to stay within clear safety boundaries.

For medical practices, the value is not only in answering more calls. The value is in answering them safely, consistently, and professionally.

Services used

• FibreMax AI Voice Agent.
• After Hours AI Assistant.
• Medical reception assistant prompt design.
• Clinic knowledge base setup.
• Emergency notice scripting.
• Structured message taking.
• Patient enquiry capture.
• Referral enquiry handling.
• Appointment enquiry message capture.
• Safe medical advice boundaries.
• Email message delivery to the clinic team.

Future phase

The first phase focuses on after hours answering and message taking.

Future phases could include.

• Separate call flows for new patients and existing patients.
• Appointment request categories for reception follow up.
• Referral enquiry capture for GP clinics and other specialists.
• Integration with practice management software, where appropriate.
• Automated routing of messages by enquiry type.
• Separate assistants for reception, after hours, appointment setting, and patient care.
• More detailed reporting on enquiry types and call reasons.

This allows a medical practice to start safely with message taking first, then expand only when the right workflows and approvals are in place.