It never guesses
Below the line you set, it stops and hands over. A confident wrong answer sent to a patient is the one failure that actually matters.
LOADING
This is the screen a practice opens on a Tuesday morning. Seven enquiries have already arrived. One of them needs a person.
Click through it. Then go to Controls and move the line — that one slider is the whole argument, and you set it, not us.
Select an enquiry.
Anyone can show software answering a question. The reason practices trust this is that it knows when not to — and that when it stops, it writes nothing, assumes nothing, and hands over everything a person needs.
Below the line you set, it stops and hands over. A confident wrong answer sent to a patient is the one failure that actually matters.
The front desk spent the morning on the person who needed them, instead of typing three appointments into a calendar.
Every message, booking and handover is recorded. If you want to check what happened on a given morning, you can.
The console shows the outcomes. These are the conversations behind them — a call before opening, a patient who did not have to explain himself twice, and the message it refused to answer.
Press play.
It behaves the same way every time, because it is scripted. We have not put a live system on this page for two reasons: real patient conversations are private, and a demo that can be steered by whoever is watching does not prove anything about how it behaves at seven in the morning with a real person waiting.
What is real is the behaviour — the line, the handover, and the record of what happened. If you would like to see it running against your own enquiries, that is what the first five days are for.