No-Show Recovery Agent
Finds the bookings about to be missed, while there is still time to sell the slot
- 8hard rules
- Things it must never do, enforced in every run
- 5handover rules
- Conditions where it stops and gives the work to a person
- 9output fields
- A structured result, not a paragraph you have to re-read
- ≤5steps per run
- Bounded work: at most 4 tool calls, enforced by the runtime
Read-only by design. This agent has no path to sending, spending, publishing or committing anything. Where that limit is the product, removing it would remove the reason to trust the output.
Overview
A missed appointment is the most expensive thing a booked business owns. The practitioner is paid, the room is heated, the slot cannot be sold twice, and somebody who wanted that time was told there was nothing. Fifteen per cent no-shows is a day and a half a week given away.
The standard reminder — everybody, twenty-four hours before — is weak twice over. It goes to the ninety per cent who were always coming, which teaches them to ignore it, and twenty-four hours is far too late to sell the slot to anyone else.
Risk is knowable, and it has nothing to do with character. A booking made six weeks out is riskier than one made yesterday. A first appointment is riskier than a fifth. Early Monday is riskier than mid-afternoon Thursday. Somebody who has missed twice is the strongest signal there is. This agent ranks on that and puts the effort where it changes an outcome.
It also works the list nobody works. A cancellation arrives, the slot returns to empty, and the person who asked for exactly that time last week is never told — because telling them means twelve calls to fill one gap. It matches the freed slot to who actually wanted it and drafts the offer.
It sends nothing. What goes to a patient under a clinic's name is a clinical and a commercial decision at once, and it stays with a person.
Benefits
Slots sold instead of lost
Because the gap was known while there was still time.
Reminders that still work
Sent to the bookings at risk rather than to everybody.
The waiting list finally worked
Matched automatically instead of by twelve phone calls.
What it does
Ranks the bookings actually at risk
Lead time, first visit or fifth, day and hour, and history. A reminder to everyone trains everyone to ignore reminders.
Early enough to resell the slot
Twenty-four hours is too late for anything but a courtesy. The intervention that recovers money happens days out.
Matches a freed slot to who wanted it
The waiting list exists and nobody works it, because working it means twelve calls to fill one gap.
Offers a move before a loss
Most people who miss an appointment would have moved it. Nobody gave them an easy way to say so.
Prices the week's exposure
What the at-risk slots are worth, so the effort has a number behind it rather than a feeling.
Never blames a patient
Findings are about lead time, slot and history — never about a person's reliability, and never a note on their record.
What this changes
What this changes about your week
- It takes over
- Working out which of next week's appointments will not happen, early enough to sell the slot to somebody else.
- Done today by
- Nobody, usually — a reminder goes to everyone the day before and the gap is discovered on the day.
- How often
- Every week, and every time a cancellation lands.
- Typical manual effort
- Twelve phone calls to fill one gap, which is why the waiting list never gets worked.
Why Healthcare in particular. A missed appointment costs the same as a full one and cannot be sold twice. Fifteen per cent no-shows is a day and a half a week given away.
Runs unattended
Started by you or by an event, and it finishes on its own. Nothing waits for someone to be at a desk.
The same standard every time
The two-hundredth item is held to the bar the first one was. Consistency is the part people cannot sustain.
It cannot act on its own
No-Show Recovery Agent has no path to sending, spending or committing. That limit is why its output is safe to act on.
How it runs
This agent runs server-side through the PROMIVO runtime. Each run is logged step by step and every tool call is permission-checked before it executes.
- 1InputValidated against the agent input schema
- 2Modelclaude-sonnet-5 with automatic fallback
- 3Toolsextract_structured_data
- 4PermissionsRuns READ actions automatically; anything higher waits for approval
- 5OutputStructured result matching the output schema
- 6LoggingTokens, duration and every step recorded
Demo
Demo dataIllustrative sample output, abridged.
Input
{
"currency": "USD",
"appointments": []
}Output
{
"atRisk": [
{
"risk": "high",
"draft": "…",
"startsAt": "2026-09-14T09:00",
"reference": "AP-8812",
"interveneBy": "2026-09-09",
"whyThisRisk": "Booked forty-one days ahead, first visit, and a nine o'clock Monday. Each of those raises the risk on its own; together they are the profile that misses most often here. Nothing about the person — this is the booking's shape.",
"stillSellableIfMissed": true
}
],
"escalate": false,
"patterns": [
"Bookings made more than thirty days ahead miss at roughly three times the rate of bookings made inside a week. A shorter booking horizon for new patients would cost you less than the misses do."
],
"disclaimer": "Risk is reported for bookings, never for people. Nothing was sent, no record was written, and no individual has been scored.",
"factorsUsed": [
"Lead time",
"First visit",
"Day and hour",
"Prior misses"
],
"slotMatches": [
{
"slot": "2026-09-11T16:30 — Nadia, laser",
"draft": "…",
"offerTo": [
"WL-233",
"WL-241"
],
"whyThese": "Both asked for a Friday afternoon with Nadia specifically in the last fortnight. Neither is at the top of the queue, which is why the list never reached them."
}
],
"escalationReason": "",
"exposureThisPeriod": 4260,
"factorsUnavailable": [
"Reminder responsiveness — no record of which reminders were opened or replied to, so it was not used."
]
}Requirements
Integrations
No integrations required.
Before you start
- Upcoming appointments with the date each was booked
- Attendance history — attended, cancelled, missed
- Slot value or service price
- Your waiting list or unmet requests, if you keep one
- Requires the Starter plan or higher
Use cases
The week ahead
Which bookings are at risk and what to do about each.
A cancellation just came in
Who asked for that exact slot and what to send them.
Setting a deposit policy
Which bookings justify one, on evidence rather than blanket.
Understanding a bad month
Whether the misses share a cause you can remove.
Pricing
Starter
Billed monthly through your PROMIVO subscription. Cancel at any time.
Runs consume your plan allowance for agent executions and tokens. See plan limits
FAQ
Does it message patients?
No. It drafts and ranks; sending stays with your team. A message to a patient under your name carries consent, preference and clinical rules that belong to you.
Does it score patients on reliability?
No, and it will not. It reports risk for a booking — lead time, slot, visit number, prior misses — never a rating of a person, and it writes nothing to anyone's record. A patient reliability score gets used to refuse care, which is a different product with a different ethics.
Does it see clinical information?
It should not be given any. It works from appointment metadata: when it was booked, when it is, which service, whether the person attended before. It never asks for or repeats a reason for a visit.
How much history does it need?
It can rank on lead time and slot alone from the first day. Prior-attendance patterns need a few months, and it states which of its factors your data actually supports rather than implying all of them.
Reviews
Reviews open once customers have run this agent.
Need this agent to work differently?
Tell us what to change and our team will scope a customised version for your business.
Customize this agent