AI Reception: When a Patient Asks for a Person
“Can I speak to someone?” is a simple request. For a patient who is already having difficulty, the answer matters more than how natural the automated voice sounds.
When reviewing an AI receptionist, follow that request all the way to its outcome. A message saying that a transfer has started does not tell you whether the patient reached the practice team or understood what would happen next.
Follow the handover to its destination
Ask the supplier to demonstrate an ordinary request for human assistance. Check the destination with the colleagues who will receive the call. They should know why the patient is arriving and what information, if any, accompanies the handover.
Then review the same journey when the receiving team is busy. A successful demonstration during a quiet period leaves this question open. Establish what the patient hears, what options are available and who owns any follow-up.
Look at the experience from both sides
The patient needs a clear route. The receptionist needs a manageable arrival. Review these together rather than treating the transfer as a purely technical event.
| Patient question | Practice question |
|---|---|
| How do I ask for help? | Where does that request arrive? |
| Am I waiting for someone? | Can staff see a waiting or unfinished request? |
| Will I need to explain it again? | What information is available to the receiving colleague? |
| What happens if nobody answers? | Who is responsible for the next action? |
Include people who may need extra help
A patient may struggle with the connection, need something repeated or find an automated conversation difficult. Include those experiences in the review. Ask reception colleagues which situations they would want handled differently and make sure the proposed route reflects the practice’s needs.
Use fictional details when demonstrating the process. The aim is to understand the journey, without introducing real patient information into a supplier comparison.
Measure completed handovers
Record transfers requested, transfers attempted and patients who actually reached the intended team. Keep abandoned or incomplete journeys visible. These measures answer different questions and should not be combined into a single headline without explanation.
For a pilot, agree how staff will report difficult handovers and how the practice will review them. A brief review of several complete journeys can reveal more useful information than a dashboard number on its own.
Review EMMA against your practice’s needs
EMMA is QuantumLoopAI’s AI patient access system for NHS GP surgeries and operates during surgery hours. Bring your handover requirements to a demonstration, including the arrangements for a busy reception team. Clinical decisions remain with appropriately qualified staff.
Use our supplier evaluation guide to structure the discussion, or book an EMMA demonstration.