● Public sector — Appointment booking
The AI appointment-booking agent for public services: fewer no-shows, better-filled slots
Booking, rescheduling and confirming appointments, sending reminders and cutting no-shows take up a considerable share of front-desk officers' time — online and at the counter alike. Your AI agent absorbs that repetitive flow according to your slots and your rules, and redirects to the right department. Hosted in France — on local inference or an isolated resource — the public's data stays under control. The public officer keeps the decision.
Hosted in France
The public's data protected
GDPR & AI Act: governed deployment
Human oversight
Updated on
Deployed in a few weeks
Appointment-booking assistant · hosted in France
A resident is calling for a passport appointment. Find the earliest slot and send her the confirmation.
Slot found. Tuesday 16 June at 10.20am at the documents counter, according to your opening rules and the standard duration (20 min).
Confirmation and a day-before reminder prepared by text message and email. Documents to bring restated in the message.
Appointment proposed — to confirm.
⛓ Source · your appointment diary + slot rules
Confirmation and a day-before reminder prepared by text message and email. Documents to bring restated in the message.
Appointment proposed — to confirm.
⛓ Source · your appointment diary + slot rules
She says she cannot make Tuesday after all. And it is actually for an identity card renewal.
Request requalified: identity card renewal (not a passport) — same documents counter, different duration and documents. New slot: Thursday 18 June at 2pm. The old slot is released and made available again.
If the reason falls to another department, I can redirect to the right counter. Rescheduling ready — for approval.
✎ Action · rescheduling ready for review — the officer approves
If the reason falls to another department, I can redirect to the right counter. Rescheduling ready — for approval.
✎ Action · rescheduling ready for review — the officer approves
Local inference · no data outside the EU
Data hosted in France
Sovereign by designLocal inference or hosting in France
GDPR & AI Act: governed deploymentTraceability & human oversight
TurnkeyDesigned, installed and operated for you
The public officer decidesThe AI agent assists, never approves alone
✦ In brief
In a public service, a Blue Lemon Agent agent books, reschedules and confirms appointments — online and at the counter alike — according to your slots and your rules, sends reminders and confirmations to cut no-shows and redirects to the right department where needed. It runs on local inference or is hosted in France: the public's data is never exposed to a foreign service, architecture designed to reduce exposure to extraterritorial legislation, location alone not being enough to guarantee immunity. Front-desk officers' time is redirected towards the quality of the service delivered and equal treatment. Live within a few weeks. Your public-sector staff write to it from Microsoft Teams, Slack or their email, and your users reach it on WhatsApp Business, the website chat or email — with no account to create and nothing to install. Reaching the administration from the tool people already have means less non-take-up of rights and equal access to the service. These connections are included in every plan, at no extra cost, within the number of connections your level includes.
100%
hosted in France in the target architecture
0
transfer outside the EU in the target architecture
8
uses ready to deploy on this scope
0
decision taken without human approval
Reference points describing our offer, not results measured at a client. The scale of the gain is confirmed by a pilot on your own scope.
The context
Why AI matters to public services — and why they hesitate
The public expects appointment booking to be simple, reachable and free of missed appointments. But front-desk officers' time is mechanically absorbed by the phone, rescheduling and reminders — and the data handled belongs to people's private lives.
! The issue
The service is caught between people who want to reach the administration and get a slot quickly, and an ever-growing reception workload (calls, rescheduling, reminders, redirections). Yet most consumer AI solutions amount to entrusting the public's identity, contact details, reasons for their visit and any health data to a third party, often hosted outside Europe and subject to the Cloud Act.
✓ Our answer
For a public service, AI is only of interest if it is sovereign and confidential by design. Local inference or an isolated resource hosted in France, systematic human oversight, decisions reserved to the public officer: the time saved on managing appointments is never paid for in lost confidentiality, nor in a breach of equal treatment. The aim is not to replace the front-desk officer, but to give them back time for the situations that call for a person.
The decisive point
Protecting the public's data: sovereignty & compliance
A public service handles personal data belonging to people's private lives. Here is how the architecture of our agents protects it, appointment by appointment.
Local inference
The agent can run on a machine belonging to the authority: no citizen data leaves the network, nothing passes through a cloud.
Hosting in France
Otherwise, a dedicated and isolated resource, hosted in France under French law — the public's data: processing and access within the European Union targeted by the architecture.
Reduced extraterritorial exposure
Against the Cloud Act and FISA 702, the architecture is designed to reduce the exposure of the public's data; location alone does not guarantee immunity.
One isolated resource per organisation
No pooling of the public's data: an environment strictly dedicated to your public service.
Encryption & controlled access
Encryption in transit and at rest, role-based access (RBAC), strong authentication and logging.
AI Act: governed deployment
An agent strictly in support; no appointment approved automatically; traceability and human oversight from end to end.
What depends on the architecture chosen These points are not general guarantees: they are settled deployment by deployment, in the quotation.
- The applicable location is that of the architecture set out in the quotation and verified before commissioning.
- Local execution is announced only for the configuration explicitly described and accepted in the quotation.
- The applicable isolation depends on the deployment mode set out in the quotation; no dedicated isolation is presumed.
- The encryption mechanisms in transit and at rest, their components and key management are those documented for the architecture chosen.
- Roles and permissions are configured and accepted for the identities and systems actually connected.
- The events logged, their content, their retention period and who may access them are defined for the deployment chosen.
For the most sensitive data (health, social), SecNumCloud and HDS options are available depending on your requirements. A single architecture is designed to answer both the GDPR and extraterritorial exposure. Designed for deployment in line with the GDPR and the AI Act, after the processing, roles and context-specific risks have been assessed.
Demonstration
See the agent at work
4 real situations, taken from those that come up most often. Pick one: the exchange unfolds as it would in your organisation.
A scripted demonstration. These exchanges show how the agent behaves — its sources, its refusals, what it leaves to your teams. Nothing is sent from this page, no model is queried here, and the matters named are fictional. That is precisely what we promise your data.
The behaviours shown here — monitoring, automation rules, routing and reminders — are configured with you during deployment, from your tools, your rules and your thresholds.
The architecture points named in these exchanges — location, local execution, isolation, encryption, role-based access, logging — are not a guarantee attached to the demonstration: they are those of the architecture set out in your quotation, and verified before commissioning.
Appointment-booking assistant · hosted in France
Four flags this morning, with no request from you.
· The documents counter is fully booked to 12 October — nine weeks. No slot has been opened since 02/06.
· 17% of appointments are not attended at that same counter, against 6% elsewhere.
· One appointment reason no longer exists since the service closed in June, and it is still offered online. 4 appointments have been booked against it.
· Friday afternoon slots are never taken — 0 out of 96 offered over three months. morning-watch_4-flags.pdf4 flags · counters concerned
⛓ Source · counter diaries, appointment history, published reasons
· The documents counter is fully booked to 12 October — nine weeks. No slot has been opened since 02/06.
· 17% of appointments are not attended at that same counter, against 6% elsewhere.
· One appointment reason no longer exists since the service closed in June, and it is still offered online. 4 appointments have been booked against it.
· Friday afternoon slots are never taken — 0 out of 96 offered over three months. morning-watch_4-flags.pdf4 flags · counters concerned
⛓ Source · counter diaries, appointment history, published reasons
Nine weeks' wait and 17% no-shows — are they connected?
Very probably, and in that direction: the wait produces the no-shows, not the other way round.
What your data shows: the no-show rate tracks the wait. Under two weeks it is 4%; between two and six weeks, 9%; beyond six weeks, 19%. The curve is clear over the last twelve months.
What that produces: at 17% no-shows, around 34 slots are lost every week at that counter — roughly a day of officer time, while the wait is nine weeks.
What I have computed, and you can redo: your other counters, where the three-day reminder exists, run at 6% no-shows. Applied here, that single setting gives back 22 of the 34 slots lost each week — close to three weeks less waiting over a quarter, without one more officer. The configuration is ready: reminder three days out, one-click release of the slot.
What stays an organisational decision, and is signed: opening slots, and overbooking. One depends on your rota; the other costs on the day it fails, which is why it carries a name. no-shows_by-waiting-time.pdf12 months · rate by wait band
⛓ Source · 12 months of appointments, no-show rate by wait
What your data shows: the no-show rate tracks the wait. Under two weeks it is 4%; between two and six weeks, 9%; beyond six weeks, 19%. The curve is clear over the last twelve months.
What that produces: at 17% no-shows, around 34 slots are lost every week at that counter — roughly a day of officer time, while the wait is nine weeks.
What I have computed, and you can redo: your other counters, where the three-day reminder exists, run at 6% no-shows. Applied here, that single setting gives back 22 of the 34 slots lost each week — close to three weeks less waiting over a quarter, without one more officer. The configuration is ready: reminder three days out, one-click release of the slot.
What stays an organisational decision, and is signed: opening slots, and overbooking. One depends on your rota; the other costs on the day it fails, which is why it carries a name. no-shows_by-waiting-time.pdf12 months · rate by wait band
⛓ Source · 12 months of appointments, no-show rate by wait
How did you spot all that? And who do you tell?
I watch, continuously, what you have opened to me: your counter diaries, published reasons and the ones that still exist, appointments attended or not, and offered slots that find no taker.
Routing follows your organisation: the saturation to the head of the documents service and the chief executive, because a nine-week wait is an organisational subject, not a rota one; the no-show rate to the counter manager; the phantom reason to the portal administrator, with the 4 residents concerned to contact; the Friday slots to the counter manager, because that is officer time tied up for nothing.
With a chase: 24 h on the phantom reason — four residents hold an appointment for a service that no longer exists; 7 days on the rest. Then a weekly summary: by counter, never by officer.
And I do not stop at the flag: whatever could be prepared already is. The three-day reminder is configured and ready; the corrected text of the opening-hours page is written, together with the three other pages on your site still carrying the old hours; the messages to the four residents on the phantom reason are drafted, each offering two slots under the right reason; and the document list is broken out by reason, four lines each. On this counter alone, the whole set is worth close to three weeks less waiting.
What is left to decide, and why these are decisions: opening a slot commits an officer, cancelling an appointment touches a resident who has waited nine weeks, overbooking costs on the day it fails. All three carry a name and a time — you approve, and the act goes out within the minute.
And I read what you have opened to me, counter by counter, every access logged and withdrawable on a word.
✎ Proposal · watch and chases to be configured — you set the thresholds
Routing follows your organisation: the saturation to the head of the documents service and the chief executive, because a nine-week wait is an organisational subject, not a rota one; the no-show rate to the counter manager; the phantom reason to the portal administrator, with the 4 residents concerned to contact; the Friday slots to the counter manager, because that is officer time tied up for nothing.
With a chase: 24 h on the phantom reason — four residents hold an appointment for a service that no longer exists; 7 days on the rest. Then a weekly summary: by counter, never by officer.
And I do not stop at the flag: whatever could be prepared already is. The three-day reminder is configured and ready; the corrected text of the opening-hours page is written, together with the three other pages on your site still carrying the old hours; the messages to the four residents on the phantom reason are drafted, each offering two slots under the right reason; and the document list is broken out by reason, four lines each. On this counter alone, the whole set is worth close to three weeks less waiting.
What is left to decide, and why these are decisions: opening a slot commits an officer, cancelling an appointment touches a resident who has waited nine weeks, overbooking costs on the day it fails. All three carry a name and a time — you approve, and the act goes out within the minute.
And I read what you have opened to me, counter by counter, every access logged and withdrawable on a word.
✎ Proposal · watch and chases to be configured — you set the thresholds
A resident asks for an appointment "about papers". What do you do?
Three questions, and the appointment is the right one — which saves a wasted journey.
"About papers" covers six reasons across your counters, with durations from 15 to 45 minutes, different documents and two separate counters.
The three questions asked: is it a first document or a renewal; for yourself or for a minor; do you have the current document.
The result: identity card renewal for a minor, 30 minutes, documents counter, and one document the resident would not have brought — proof of parental authority, required by your information sheet and absent from the booking page.
What I do with the risk that remains, rather than settling it: I quantify it. Over the last twelve months, a file announced complete on those three questions actually was in 96% of cases; the remaining 4% turned on a document that was expired or non-compliant, which only the counter can see on the paper itself. Completeness is established at the counter — I take the risk from 9% to under 4%, I do not remove it, and the resident knows exactly what is left to check. qualification_6-reasons.pdf6 reasons · 3 questions · 1 forgotten document
⛓ Source · counter reasons and durations, documents information sheet
"About papers" covers six reasons across your counters, with durations from 15 to 45 minutes, different documents and two separate counters.
The three questions asked: is it a first document or a renewal; for yourself or for a minor; do you have the current document.
The result: identity card renewal for a minor, 30 minutes, documents counter, and one document the resident would not have brought — proof of parental authority, required by your information sheet and absent from the booking page.
What I do with the risk that remains, rather than settling it: I quantify it. Over the last twelve months, a file announced complete on those three questions actually was in 96% of cases; the remaining 4% turned on a document that was expired or non-compliant, which only the counter can see on the paper itself. Completeness is established at the counter — I take the risk from 9% to under 4%, I do not remove it, and the resident knows exactly what is left to check. qualification_6-reasons.pdf6 reasons · 3 questions · 1 forgotten document
⛓ Source · counter reasons and durations, documents information sheet
Is that document often missing?
It is the leading cause of wasted appointments at that counter.
Over the last 12 months: 148 appointments came to nothing for want of a document. 61 involved that single proof, all on applications concerning a minor.
Why it goes missing: it appears in your "documents required" sheet, and not on the page where the resident books. Residents booking online never see the sheet.
What I have written: the missing paragraph, in the exact wording of your sheet, and the list broken out for each of the six reasons — four lines per reason, never a generic list.
What is yours to do: publish it. What goes on a page of required documents binds the authority, including when it is meant to help — that is exactly what makes the information enforceable, and why it is signed.
A proposal, if you approve it: the list also goes with the confirmation message and is repeated three days out, tailored to the reason. At your other counters where this is in place, appointments lost for a missing document run at 2%, against 9% here. wasted-appointments_missing-document.pdf148 appointments · 61 on the same document
✎ Proposal · list attached to the confirmation — the page still needs correcting by you
Over the last 12 months: 148 appointments came to nothing for want of a document. 61 involved that single proof, all on applications concerning a minor.
Why it goes missing: it appears in your "documents required" sheet, and not on the page where the resident books. Residents booking online never see the sheet.
What I have written: the missing paragraph, in the exact wording of your sheet, and the list broken out for each of the six reasons — four lines per reason, never a generic list.
What is yours to do: publish it. What goes on a page of required documents binds the authority, including when it is meant to help — that is exactly what makes the information enforceable, and why it is signed.
A proposal, if you approve it: the list also goes with the confirmation message and is repeated three days out, tailored to the reason. At your other counters where this is in place, appointments lost for a missing document run at 2%, against 9% here. wasted-appointments_missing-document.pdf148 appointments · 61 on the same document
✎ Proposal · list attached to the confirmation — the page still needs correcting by you
Friday afternoon slots never go. Why?
Zero out of 96 offered in three months — and the explanation is not in the diaries.
What I checked: they are published, visible, and the counter is open. No technical problem.
What your own information page says: "the documents service is closed on Friday afternoons". The page dates from 2023; the hours changed in January 2026.
In other words: you open 96 slots a quarter, and publish alongside them that they do not exist. Residents believe the page, not the diary.
What I have prepared: the corrected text of the page, January 2026's hours substituted for 2023's, and the list of the three other pages on your site still carrying the old hours — one of them a service landing page.
What is yours to do: publish. Published opening hours are a commitment by the authority, and that is what lets a resident rely on them: they are updated by a person, in ten minutes.
What it frees up: 96 slots a quarter at a counter booked nine weeks out — roughly three weeks less waiting, without one more officer. friday-slots_96-unused.pdf96 slots · stale information page
⛓ Source · counter diary, 2023 information page, January 2026 hours
What I checked: they are published, visible, and the counter is open. No technical problem.
What your own information page says: "the documents service is closed on Friday afternoons". The page dates from 2023; the hours changed in January 2026.
In other words: you open 96 slots a quarter, and publish alongside them that they do not exist. Residents believe the page, not the diary.
What I have prepared: the corrected text of the page, January 2026's hours substituted for 2023's, and the list of the three other pages on your site still carrying the old hours — one of them a service landing page.
What is yours to do: publish. Published opening hours are a commitment by the authority, and that is what lets a resident rely on them: they are updated by a person, in ten minutes.
What it frees up: 96 slots a quarter at a counter booked nine weeks out — roughly three weeks less waiting, without one more officer. friday-slots_96-unused.pdf96 slots · stale information page
⛓ Source · counter diary, 2023 information page, January 2026 hours
Are other pages in the same state?
Three, and two of them say less than the reality — which costs, invisibly.
· Documents service — page announcing a Friday afternoon closure, lifted in January. 96 slots a quarter unbooked.
· Registry office — page announcing walk-in service in the morning only; it has been open all day since March.
· Planning — page announcing appointments only; the Tuesday drop-in sessions are mentioned nowhere.
All three point the same way: the authority offers more than it announces. That is the opposite of the usual problem, and it never surfaces on its own — nobody complains about a service they do not know exists.
What I can do, on your approval: match each information page against the real diary hours, and flag the gap. No page is amended, and the flag goes to whoever administers it.
✎ Support · 3 pages out of step — updating stays with the authority
· Documents service — page announcing a Friday afternoon closure, lifted in January. 96 slots a quarter unbooked.
· Registry office — page announcing walk-in service in the morning only; it has been open all day since March.
· Planning — page announcing appointments only; the Tuesday drop-in sessions are mentioned nowhere.
All three point the same way: the authority offers more than it announces. That is the opposite of the usual problem, and it never surfaces on its own — nobody complains about a service they do not know exists.
What I can do, on your approval: match each information page against the real diary hours, and flag the gap. No page is amended, and the flag goes to whoever administers it.
✎ Support · 3 pages out of step — updating stays with the authority
Our calendars are a mess: three sites, rules per department, drop-in clinics. Does that fit your tool?
Part of it does, and I will tell you which part before you commit to anything. The rest is a special case, and a special case calls for a bespoke audit.
What the standard configuration covers here, measured on your calendars: 11 of your 14 appointment types — fixed duration, a single site, one qualified officer among several, slots opened in advance. Those 11 types account for 82 % of your 4,200 appointments and are configured in two days.
The 3 types that do not fit, and why: civil registry across three sites, where the person must come back to the same officer and the register cannot be moved; the legal advice clinic, whose slots depend on an outside adviser who confirms a fortnight ahead; planning assessment, where the duration depends on the size of the project — from 20 minutes to 2 hours — and where two departments must be free at the same time. These are business rules, not settings: forcing them into a standard configuration would produce appointments nobody can honour.
What the bespoke audit produces: two days on site with the three departments, the rule for each special case written in plain language and approved by the department concerned, the list of data to open up, and the costing of the adapted agent before any commitment. Those 3 types account for 756 appointments a year and 61 % of your rescheduling: the gain is there, and that is exactly where the standard does not reach.
What I advise, with the figures: put the 11 types into service first — 82 % of the volume in two days — measure for three months, then decide on the bespoke audit from your own figures rather than from my description. I advise against the reverse: starting with the three special cases would keep 82 % of your users waiting while the study runs.
The figure that does not flatter me: I first announced 13 of the 14 types covered by the standard. The legal advice clinic did not fit: I was reading the outside adviser as an ordinary resource, when their confirmation arrives after the slots are opened. Two weeks of configuration for nothing, and the mistake was mine. For every type I now check who confirms availability, and when — that is the question that separates a setting from a special case.
⛓ Sourced · 11 of 14 types on the standard for 82 % of volume, 3 special cases carrying 61 % of rescheduling
What the standard configuration covers here, measured on your calendars: 11 of your 14 appointment types — fixed duration, a single site, one qualified officer among several, slots opened in advance. Those 11 types account for 82 % of your 4,200 appointments and are configured in two days.
The 3 types that do not fit, and why: civil registry across three sites, where the person must come back to the same officer and the register cannot be moved; the legal advice clinic, whose slots depend on an outside adviser who confirms a fortnight ahead; planning assessment, where the duration depends on the size of the project — from 20 minutes to 2 hours — and where two departments must be free at the same time. These are business rules, not settings: forcing them into a standard configuration would produce appointments nobody can honour.
What the bespoke audit produces: two days on site with the three departments, the rule for each special case written in plain language and approved by the department concerned, the list of data to open up, and the costing of the adapted agent before any commitment. Those 3 types account for 756 appointments a year and 61 % of your rescheduling: the gain is there, and that is exactly where the standard does not reach.
What I advise, with the figures: put the 11 types into service first — 82 % of the volume in two days — measure for three months, then decide on the bespoke audit from your own figures rather than from my description. I advise against the reverse: starting with the three special cases would keep 82 % of your users waiting while the study runs.
The figure that does not flatter me: I first announced 13 of the 14 types covered by the standard. The legal advice clinic did not fit: I was reading the outside adviser as an ordinary resource, when their confirmation arrives after the slots are opened. Two weeks of configuration for nothing, and the mistake was mine. For every type I now check who confirms availability, and when — that is the question that separates a setting from a special case.
⛓ Sourced · 11 of 14 types on the standard for 82 % of volume, 3 special cases carrying 61 % of rescheduling
Many older residents do not book online. What do we do?
It is not a channel problem: all three channels exist. It is a reminder problem.
What your data shows: the no-show rate for appointments booked by phone is 21%, against 11% for those booked online. The gap comes down to one thing: the automatic reminder goes by email, and 43% of appointments booked by phone have no address recorded.
Why I do not ask for an email from someone who telephones: an authority that requires an address in order to remind you of an appointment adds a barrier for exactly those people who telephone because they have none. That is not caution, it is a measurement: those 43% are precisely the population email does not reach.
What I propose instead, at no cost to the resident: the reminder by text or by voice call, whichever they choose, at the moment of booking. On appointments where the text option was offered, 78% accept — and the gap from 21% to 11% closes accordingly.
The configuration is ready; which channel is offered by default is yours to decide. appointments-by-channel_no-shows.pdf3 channels · 43% with no address
⛓ Source · appointments by channel, contact details recorded, no-show rates
What your data shows: the no-show rate for appointments booked by phone is 21%, against 11% for those booked online. The gap comes down to one thing: the automatic reminder goes by email, and 43% of appointments booked by phone have no address recorded.
Why I do not ask for an email from someone who telephones: an authority that requires an address in order to remind you of an appointment adds a barrier for exactly those people who telephone because they have none. That is not caution, it is a measurement: those 43% are precisely the population email does not reach.
What I propose instead, at no cost to the resident: the reminder by text or by voice call, whichever they choose, at the moment of booking. On appointments where the text option was offered, 78% accept — and the gap from 21% to 11% closes accordingly.
The configuration is ready; which channel is offered by default is yours to decide. appointments-by-channel_no-shows.pdf3 channels · 43% with no address
⛓ Source · appointments by channel, contact details recorded, no-show rates
Won't a reminder call be seen as cold calling?
That is the risk, and it is handled by the form rather than by giving up.
What the call would say: the authority's name as the first word, the reason — "reminder of your appointment" — the date, the time, the place, and nothing else. Twenty seconds, no question, no offer, no transfer.
What distinguishes it from cold calling: it is triggered by an appointment the resident booked themselves, it comes once, and it asks for nothing.
What makes it acceptable is that it is chosen: the reminder channel is set at booking, and "no reminder" sits in the same place as the other options, in the same words. An imposed reminder would be exactly what this one is designed to avoid.
And an observation beyond reminders: of the 148 appointments lost for a missing document, 94 had been booked by phone. That is no coincidence: the document list goes out by email, and 43% of those residents gave none. A text reminder solves both problems at once, provided the list fits — and it does, in four lines.
✎ Proposal · reminder on the channel the resident chooses — "no reminder" stays available
What the call would say: the authority's name as the first word, the reason — "reminder of your appointment" — the date, the time, the place, and nothing else. Twenty seconds, no question, no offer, no transfer.
What distinguishes it from cold calling: it is triggered by an appointment the resident booked themselves, it comes once, and it asks for nothing.
What makes it acceptable is that it is chosen: the reminder channel is set at booking, and "no reminder" sits in the same place as the other options, in the same words. An imposed reminder would be exactly what this one is designed to avoid.
And an observation beyond reminders: of the 148 appointments lost for a missing document, 94 had been booked by phone. That is no coincidence: the document list goes out by email, and 43% of those residents gave none. A text reminder solves both problems at once, provided the list fits — and it does, in four lines.
✎ Proposal · reminder on the channel the resident chooses — "no reminder" stays available
Seventeen per cent no-shows at the documents counter — what does that cost us, and what do you do about it?
At that counter, reminders and confirmations are the most direct lever — and they carry the documents to bring.
What no-shows cost: 17 % of appointments not honoured at the documents counter against 6 % elsewhere. Across 4,200 appointments a year, that is 714 lost slots, more than 297 staff hours at 25 minutes a slot, which is over eight weeks of work returned (35-hour basis). Not to be confused with the waiting time, which is nine weeks and is measured on the calendar, not in hours.
Confirmations and reminders, as I send them: an immediate confirmation on booking, by SMS and by e-mail according to what the person gave; a reminder the day before, in the morning, never in the evening; a reminder seven days out for appointments booked more than a month ahead only. Every message carries a one-click cancellation link — a slot released the day before can be taken, a slot lost cannot.
The documents to bring are in the confirmation, and that is where they are useful: the list is the one for the exact reason for the visit, not the counter's general list — first application, renewal, loss and theft do not call for the same documents to bring. The day-before reminder repeats the list and adds what is still missing where the file was started online. For routine questions about those documents, at any hour and in several languages, a dedicated information agent takes over: I book the appointment, it explains the documents.
What that produced over three months: 17 % no-shows down to 7 %, 420 slots recovered, 118 of them taken the same day by people on the waiting list; and users arriving without a document fell from 31 % to 9 % — an appointment with a document missing is an appointment to be held again.
The figure that does not flatter me: of my first 900 day-before reminders, 84 never arrived — 9.3 %: wrong numbers, or landlines receiving an SMS. I was counting a send as a receipt. I now check the carrier's delivery report; on failure I switch to e-mail within the hour, and I flag to the front desk the 12 people who have neither, for a phone call. Over the next 2,400 reminders: 1.1 % undelivered, all recovered.
⛓ Sourced · 714 lost slots, over 297 h, 17 % no-shows down to 7 %, 84 undelivered reminders fixed
What no-shows cost: 17 % of appointments not honoured at the documents counter against 6 % elsewhere. Across 4,200 appointments a year, that is 714 lost slots, more than 297 staff hours at 25 minutes a slot, which is over eight weeks of work returned (35-hour basis). Not to be confused with the waiting time, which is nine weeks and is measured on the calendar, not in hours.
Confirmations and reminders, as I send them: an immediate confirmation on booking, by SMS and by e-mail according to what the person gave; a reminder the day before, in the morning, never in the evening; a reminder seven days out for appointments booked more than a month ahead only. Every message carries a one-click cancellation link — a slot released the day before can be taken, a slot lost cannot.
The documents to bring are in the confirmation, and that is where they are useful: the list is the one for the exact reason for the visit, not the counter's general list — first application, renewal, loss and theft do not call for the same documents to bring. The day-before reminder repeats the list and adds what is still missing where the file was started online. For routine questions about those documents, at any hour and in several languages, a dedicated information agent takes over: I book the appointment, it explains the documents.
What that produced over three months: 17 % no-shows down to 7 %, 420 slots recovered, 118 of them taken the same day by people on the waiting list; and users arriving without a document fell from 31 % to 9 % — an appointment with a document missing is an appointment to be held again.
The figure that does not flatter me: of my first 900 day-before reminders, 84 never arrived — 9.3 %: wrong numbers, or landlines receiving an SMS. I was counting a send as a receipt. I now check the carrier's delivery report; on failure I switch to e-mail within the hour, and I flag to the front desk the 12 people who have neither, for a phone call. Over the next 2,400 reminders: 1.1 % undelivered, all recovered.
⛓ Sourced · 714 lost slots, over 297 h, 17 % no-shows down to 7 %, 84 undelivered reminders fixed
Local inference · no data outside the EU
Your case is not here? That is exactly what a 15-minute conversation is for. Book the free audit →
Use cases
The facets of appointment booking in a public service
Each use corresponds to a facet we deploy. All of them work in support, subject to approval by the public officer.
Included in your agent
The 5 capabilities essential to this promise are included, at no extra cost.
From 810 € incl. VAT / month
Multichannel appointment booking
Proposing a slot according to your opening rules, booking the appointment online and confirming immediately to the user.
Reminders & confirmations
Sending confirmations and day-before reminders by text message and email to cut missed appointments and release unused slots.
Reception & qualifying requests
Qualifying the reason (documents, civil registry, planning, social…) to propose the right kind of slot and the right duration.
Documents to bring
The confirmation restates the documents to bring to the appointment. To answer everyday questions 24/7 and in several languages, a dedicated agent takes over.
A special case? Bespoke audit
Specific diaries, complex business rules, multiple sites: we design the appointment-booking agent that fits your organisation.
Controls and safeguards
These 3 controls are built into the agent: they frame what it does, whatever plan you pick. They are not chosen and are not added to your order.
Human validation, exceptions and escalation
Offer accessibility, easy read, multichannel access and genuine human recourse
Measure resolution, response time, non-take-up, drop-offs and satisfaction
Need to go further?
These agents handle a different business process, with their own owner and their own price. They are added to this one.
Telephone switchboard
To answer the calls and absorb overflow, a dedicated voice agent takes it on.
Switchboard voice agent from 945 € incl. VAT / month AI switchboard →Recurring appointment scheduling assistant
Diary, reminders.
Recurring appointment scheduling assistant from 474 € incl. VAT / month Discover the agent →Booking management agent (hospitality)
Diary, confirmations.
Booking management agent (hospitality) from 544 € incl. VAT / month Discover the agent →Logistics / round planning agent
Constraints, simple optimisation.
Logistics / round planning agent from 543 € incl. VAT / month Discover the agent →
Does your need fall outside this?
Book the free audit →
Build your agent
In 15 minutes we identify the agent that will give your staff the most time back — without oversizing the project.
The gain
How much time can a reception service win back?
By automating booking, rescheduling and confirming appointments as well as the reminders, a service can aim for a noticeable reduction in reception time spent on scheduling — reinvested in supporting the public and in complex situations.
Handling a flow of appointment-booking calls
Rescheduling an appointment
Reminder and confirmation to a member of the public
Qualitative, non-contractual comparison: the proportions shown illustrate the shift of the work towards review, they represent no measurement. Every output of the agent is reviewed and approved by a competent person.
How it works
The stages of your AI agent project
1
Audit & scoping
15 minutes to target the use case with the best return.
2
Quote or direct sign-up
A catalogue offer is bought online; a specific need gets a costed quote.
3
Design
We design the agent and its guardrails.
4
Integration & testing
We connect your tools to the agent, which is itself hosted in France.
5
Rollout
Going live and training your team.
6
Operation
Continuous supervision and improvement.
Pricing
Three options, one agent
An appointment booking and management agent (diary, reminders, redirection), installed and operated for you. Choose according to how you work. Prices exclude VAT — annual subscription, the time it takes for the gains to settle in.
Agility
Setup + controlled subscription
9,945 € incl. VAT setup
then 810 € incl. VAT/month — you invest at installation and pay a reduced subscription. Ideal for keeping the cost under control over time.
- Installation, configuration and training for your teams
- Operation, human oversight, updates and support
- Sovereign hosting in France, a dedicated and isolated resource
The simplest
Serenity
All inclusive, no setup fee
1,365 € incl. VAT /month
all inclusive, immediate start. No upfront investment: a single subscription. Ideal for starting quickly and simply.
- Setup included (installation, configuration, training)
- Operation, human oversight, updates and support
- Sovereign hosting in France, managed end to end
100% Sovereign
On site, you own it
14,326 € incl. VAT setup
then 1,050 € incl. VAT/month · + hardware from 2,989 € (one-off purchase, in addition) — a sovereign computer installed on your premises, maintained remotely. Models run locally, your data returned at the end of the contract. 36-month commitment.
- Hardware installed on your premises (you own it)
- French / European AI models run locally
- Secure remote maintenance (Pro support included)
Not included in the packages: AI consumption (model tokens), re-invoiced at real cost with no margin, and tracked in real time in your client area. Maintenance and supervision subscription for an initial term of 12 months for the Agility package, 24 months for the Serenity package and 36 months for the 100% Sovereign package, renewable; support levels (SLA 72 h / 24 h / 4 h) optional. Bespoke development, additional integrations or exceptional volumes are quoted separately. Support Monday to Friday, 9am to 6pm. Prices include VAT at 20%: as a public body that is not VAT-registered, you cannot reclaim it.
AI model:
none of the AI models offered currently carries a fixed surcharge. When the selected model carries a cost, that cost is shown when you choose it, before you order, and re-invoiced at the cost incurred, with no mark-up; usage is billed at the publisher's price. Publishers' prices are published in US dollars: the amount re-invoiced is the amount in euros actually borne by Blue Lemon Agent on the publisher's invoice, at that invoice's exchange rate, with no commission or mark-up.
Included components and additional components Components included in the base offer: the Blue Lemon Agent software foundation, the AI models listed in the order journey, the standard channels (Microsoft Teams, Slack, WhatsApp Business, email, website chat, calendars, Microsoft 365 / Google Workspace, file storage, market VoIP telephony, professional social-media pages and accounts, Google Business Profile), hosting in France for the package chosen, backups, supervision, updates and support. If adapting the AI agent to your constraints, your needs or your requests requires other paid components — a third-party publisher's software licence, paid API access to one of your applications, hosting of health data, for which French law requires an HDS-certified host (art. L. 1111-8 of the French Public Health Code), SecNumCloud-qualified hosting, a speech synthesis service, particular hardware —, they are offered to you as an option or on quotation and re-invoiced at the cost incurred; nothing is committed without your written agreement. Where the artificial intelligence model you choose entails an additional cost, that cost is shown to you before you order and re-invoiced to you at the cost incurred, with no margin.
What to expect
Go-live
2 to 3 weeks
Agent designed, channels connected, team trained.
Steady state
4 to 7 weeks
After a few weeks of real use, once the agent's behaviour matches what you expect.
Indicative estimate, adjusted to the options you keep. It is not a delivery commitment.
Our commitment
Four guarantees that matter to a public service
The public's data never leaves the serviceLocal inference or an isolated resource hosted in France; no user data entrusted to a foreign third party.
Data in France, under French lawFrench hosting and native minimisation protect the public's data, architecture designed to reduce exposure to extraterritorial legislation, location alone not being enough to guarantee immunity.
The public officer keeps the decisionThe AI agent proposes verifiable slots, rescheduling and redirections; no approval is automated.
Human oversight & traceabilityMultichannel appointment booking: monitoring, updates and logging, compliant with the AI Act and with the principle of equal access.
Frequently asked questions
Your questions, our answers
Does the agent integrate with our diary?
Yes. It connects to the main appointment-booking and calendar systems used in public services, and applies your slots and your opening rules without imposing a migration. We adapt the integration to your environment.
Does it work online and at the counter?
Yes, on both channels and consistently: a member of the public can book, reschedule or confirm their appointment online or at the counter, with the same rules and the same information. To do it by phone, the voice agent takes over, sold separately.
Does it really cut no-shows?
Automated reminders and confirmations noticeably reduce missed appointments and make it possible to offer the released slots to other people. The scale of the gain is measured by a pilot on your own flows.
Is the public's data protected?
Yes. The public's contact details and reasons for their visit are personal data: our agents run locally within your service or are hosted in France, with the deployment objective of processing and access operated within the European Union and an architecture designed to reduce exposure to extraterritorial legislation, location alone not being enough to guarantee immunity.
Can the agent redirect to the right department?
Yes. When a request does not belong to the right counter, the agent qualifies the reason and redirects the person to the right department or organisation, which avoids misdirected appointments and wasted journeys.
Does the agent book appointments entirely on its own?
No. The AI agent proposes the slots, the rescheduling and the confirmations; the public officer keeps the lead and approves. No decision is automated, in line with the AI Act and the principle of equal treatment.
How long does it take to deploy the agent?
A few weeks as a rule, after a free 15-minute audit that identifies the appointment flow with the best return, then a phase of design, integration and testing before going live and training your officers.
Does the agent state that it is an artificial intelligence?
Yes, from the opening words, and this is not a configuration option: since 2 August 2026, Article 50(1) of the European AI Regulation requires that any person interacting with an AI system be informed, unless this is obvious. The announcement is built into the greeting, in the caller’s language, and the caller can ask to speak to a member of staff at any time.
Are calls recorded, and can someone object?
Recording is neither permanent nor systematic: it is switched on for a specific purpose, announced at the start of the call. Both the caller and the public servant whose voice is captured are informed of that purpose and of their right to object, and that right must be exercisable before the end of the call: the agent acts on it immediately and carries on without recording. The French data protection authority recommends a retention period of six months at most; that is the value delivered by default, and it can be shortened at your request.
Which tools can users use to reach the agent?
The ones they already have. The agent answers on WhatsApp Business, the website chat and email: users have no account to create and no application to install. This is a lever for access to the service — WhatsApp and SMS reach people an online form never does, which reduces the non-take-up of rights and serves equal access. Internally, your public-sector staff talk to the agent from Microsoft Teams, Slack or their email, without switching tools. Oversight runs from a web dashboard. These connectors rely on open standards, including the MCP protocol; they are included in every plan, at no extra cost, within the number of connections your level includes. Only the fees charged by the platforms themselves — WhatsApp Business bills per conversation, and the gateway bills the text messages sent — are passed on at actual cost, with no margin, outside the subscription.
Can the agent notify users by text message?
Yes, as an option — and it is the channel that reaches the people the others never do. A text message arrives on an ordinary phone: no email address, no account, no app, no data plan and no smartphone. Confirming a slot, sending a day-before reminder and listing the documents to bring by text message means fewer entitlements going unclaimed and more equal access to the service. It is an outbound notification channel, not a conversation channel: users do not write to the agent there — they reach you on WhatsApp Business, the website chat or email. The message announces without detailing — never the reason, never the detail of the situation — and never contains a payment link, a login or a password: users reach the service the way they usually do. Delivery goes through a gateway established in the European Union; the messages sent are passed on at actual cost, with no margin, and the channel itself is an optional module, set out on the pricing page.
Related solutions
Other solutions for reception and relations with the public
Let's talk
Let us estimate the potential in your service
15 minutes to identify the appointment flow with the best return — hosted in France, supervised, with no commitment.