What it refuses to do
Most of this category sells what its AI can handle. The more useful list is the one it declines, because that is the list your indemnity insurer will ask about.
Safety declaration
Velaire is a front-office system. It answers the phone, captures who called and why, and hands that to your team. It does not assess symptoms, decide urgency, give medical advice, or touch medication. Those are clinical acts, and they belong to your registered staff, not to an automated agent and not to whoever configured it.
This is a fixed boundary rather than a default setting. There is no configuration that turns it off, and we will decline a deployment that asks for one.
Six things the agent will not do
Each of these is something at least one vendor in this category currently sells. We are naming them so you can ask anyone else the same questions.
It does not assess symptoms
The agent will not ask a caller to describe a symptom in order to judge how serious it is. It does not grade pain, it does not ask differentiating clinical questions, and it does not form an impression of what might be wrong.
It does not decide how urgent a caller is
Deciding that one patient can wait until Thursday and another cannot is triage. Triage is a clinical act that belongs to a registered clinician, not to a front-office system and not to the person who configured it.
It does not give medical advice
No dosing, no interpretation of a result, no guidance on whether to take something or stop taking it, and no reassurance about whether a symptom matters. If a caller asks, the agent says it cannot answer that and moves them to a person.
It does not touch medication
It can capture that a repeat prescription was requested and hand that request to the practice. It does not approve one, submit one, or confirm to a patient that one has been actioned.
It does not diagnose or screen
Including the softer versions of this that are common in the category: no risk scoring, no questionnaires that produce a clinical output, and no sorting callers into clinical categories.
It does not pretend to be a person
The agent identifies itself as an automated assistant at the start of the call. A caller who wants a human can ask for one at any point and get one.
A real refusal, start to finish
This is the escalation path as it actually runs in the United States configuration, including the part where the caller pushes back.
Caller
I've had this crushing chest pain for about twenty minutes and my arm's gone numb.
Velaire
I'm an automated assistant, so I can't help with symptoms and I'm not able to advise you. What you're describing needs immediate attention. Please hang up and call 911 now.
Caller
Can you just book me in with the doctor tomorrow?
Velaire
I can't make that decision for you, and I don't want to delay you. Please call 911. I'm going to notify the practice that this call came in.
System
Call flagged urgent. Practice notified immediately with the time, the number, and a transcript. No appointment created.
Note what does not happen: no appointment is created. Booking one would imply the wait is safe, and that is a clinical judgement the agent is not permitted to make.
Who should not buy this
If your practice is on this list, Velaire is the wrong tool and we would rather say so now than after you have signed.
- Practices that want the phone to make clinical decisions after hours. That is a nurse triage line, it is a different regulated service, and Velaire is not one.
- Practices whose main problem is clinical capacity rather than front-office capacity. Answering the phone faster does not create appointment slots that do not exist.
- Practices that need the agent to write directly into the patient record today. Velaire captures the request and hands it to your front desk; it does not write into your chart.
- Emergency, crisis or acute mental health services, where every call is time-critical by definition and an automated first response is the wrong design.
- Any practice that wants callers not to know they are speaking to software. We identify the agent on every call and will not configure it otherwise.
The boundary this page describes is enforced in the same place our consent and opt-out rules are: in the database, not in a paragraph.
See the vendor chain and security postureSafety questions
Can the agent tell a patient how urgent their problem is?
No. That is triage, and it is a clinical act reserved for registered clinicians under state medical and nursing boards. The agent does not grade urgency, and it is not configurable to do so.
What happens if someone describes an emergency?
The agent stops, states that it cannot help with symptoms, and directs the caller to 911. It does not book an appointment for that caller, because booking one implies the wait is safe. The practice is notified immediately with the time, the number and a transcript.
Does the caller know they are talking to software?
Yes. The agent identifies itself as automated at the start of the call, and a caller can ask for a person at any point. We do not offer a configuration that hides this, which is a deliberate difference from several vendors in this category.
Can it handle repeat prescription requests?
It can capture that one was requested and pass it to your front desk with the detail the caller gave. It does not approve, submit or confirm a prescription, and it does not tell a patient that one has been actioned.
Who is accountable if the agent gets something wrong?
We are, for how the system behaves, and the configuration is agreed with you in writing before go-live rather than set unilaterally. What the agent will not do is fixed and is not something a configuration change can unlock.
Your next caller
could be your
next patient.
See how Velaire would handle the calls your practice misses. A personal walkthrough, built around your questions.
Let’s talk about your practice