+33 (0)1 87 66 00 65 · Monday to Friday, 9am–6pm Free audit (15 min)
An orchestration agent, scoped with you before it is priced. This agent coordinates several specialised agents. Its orchestration follows your actual workflows — which is why it is scoped with you rather than bought off the shelf. We establish the scope together, then the quotation commits it. The journeys described below form the scope that this review refines and the quotation commits. The specialised agents it coordinates can be ordered today. Request a quote
● Public sector — Phone reception

The AI switchboard operator for public service: no call left unanswered

At peak times, part of the public's calls go unanswered: engaged line, overflow, closing time. That is a breach of equal access to public service. Your voice agent picks up, gives information on opening hours and procedures, routes to the right department and books appointments. Hosted in France — on local inference or an isolated resource. The AI agent assists, the public officer decides.

Hosted in France No call left unanswered GDPR & AI Act: governed deployment Human oversight

Updated on

Deployed in a few weeks
Switchboard voice agent · hosted in France
It is Tuesday, a peak day on the switchboard. Take the incoming calls, give information on opening hours and procedures, and route to the right department.
On the line. Of the last 12 calls: 7 information requests (opening hours, documents for an identity card, market day) handled at the first level, 3 routed to the civil registry and planning, 2 complex requests transferred to an officer.
No call was left unanswered — the queue has cleared.
⛓ Source · the authority's knowledge base + opening hours
A resident wants an appointment about a planning file. Can you schedule it?
Request qualified: submitting a prior works declaration. Slot proposed: Thursday 11am with the planning department, with the list of documents to bring sent by text message. The request is recorded in the department's diary.
If the resident insists on speaking to a person, I transfer immediately.
✎ Action · appointment proposed — the front-desk officer confirms
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 rules alone
✦ In brief

In a local authority or a public body, a Blue Lemon Agent agent handles phone reception: it picks up, gives information on opening hours, procedures and contact details, routes to the right department and books appointments — especially at peak times and on overflow. The caller is told they are speaking to a synthesised voice; complex requests are transferred to a human officer. It runs on local inference or is hosted in France, architecture designed to reduce exposure to extraterritorial legislation, location alone not being enough to guarantee immunity. Continuity of service, equal access, and time given back to the front-desk officers. 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, email or VoIP telephony — 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
9
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 phone reception — and why authorities hesitate

The public expects a quick answer and equal access to public service, at any hour. But the switchboard is saturated at peaks, front-desk officers are constantly interrupted, and callers' data is sensitive public data.

! The issue

The switchboard is caught between people who want an immediate answer and front-desk officers overwhelmed at peak times: engaged lines, lost calls, queues. That is a breach of equal access to public service. Yet most consumer voice AI solutions amount to entrusting the public's contact details and the content of their requests to a third party, often hosted outside Europe and subject to the Cloud Act.

Our answer

For an administration, voice AI is only of interest if it is sovereign and confidential by design. Local inference or an isolated resource hosted in France, systematic transfer to a human officer on sensitive requests, clear notice to the caller: continuity of service is never paid for in exposed public data. The aim is not to replace the front-desk officer, but to absorb the peaks and give them back time for the people at the counter.

The decisive point

The public's data: sovereignty & compliance

A switchboard handles personal data belonging to users of public service. Here is how the architecture of our agents protects it, call by call.

Local inference

The agent can run on a machine belonging to the authority: no call data leaves the network, nothing passes through a public 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

Architecture designed to reduce exposure to extraterritorial legislation, location alone not being enough to guarantee immunity for the public's data: our architecture stays out of the American hyperscalers' reach, even when hosted in Europe.

One isolated resource per administration

No pooling of data: an environment strictly dedicated to your authority or institution.

Encryption & controlled access

Encryption in transit and at rest, role-based access (RBAC), strong authentication and logging of the calls handled.

AI Act: governed deployment

An agent strictly in support; the caller is told they are speaking to an AI; human transfer and 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, 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

5 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.

The public body in this demonstration

Fictional public body

Town of Naubourg — fictional local authority of 12,400 residents

Sector
Mid-sized town and its municipal social action centre — civil status, planning, schools and after-school care, social action, roads; front desk open 32 hours a week, switchboard shared with 5 neighbouring rural municipalities
Headcount
148 staff, of whom 3 on the telephone front desk — two full-time and one part-time — and one head of reception and service-user relations
Public served
12,400 residents, 6,200 households — 18,600 calls a year, 9,400 counter visits, 4,300 e-mails
Order of magnitude
12,200 of the 18,600 calls concern 15 recurring topics, 1,680 appointments booked a year, 2,900 calls taken on behalf of the 5 neighbouring municipalities
Tools in place
The town's telephone system — numbers unchanged —, departmental calendars, website, front-desk knowledge base and service-user relations software: the agent plugs into them read-only, nothing is replaced and nothing is migrated
Who decides
The head of reception and service-user relations approves the standard answers and opens the appointment slots; each department head settles what belongs to them; the chief executive signs the direct-answer mandate
Room for improvement
4,650 calls out of 18,600 go unanswered at peak hours — 25 %; the average Monday morning wait reaches 3 minutes; 78 service users out of 420 turn up for an appointment without the documents requested; and 6 of the 15 answers posted at the front desk no longer match the organisation in force

At Naubourg, three front-desk officers handle 18,600 calls a year — those of the town, those of the municipal social action centre and those of five neighbouring rural municipalities. The quality of the answers is not in question: it is availability that is missing, and a quarter of service users hang up without having spoken to anyone. The agent runs on local inference on a machine belonging to the authority, plugs into the telephone system without changing a single number, and states from its first sentence that it is a digital assistant. The exchanges below span a year, from the review of the calls to the report presented to the town council.

This public body, its figures and the exchanges that follow were invented for the demonstration. They illustrate a common situation; they describe no real service.

Switchboard voice agent · hosted in France
I have reread your last twelve months of switchboard traffic — 18,600 calls, 9,400 counter visits, 4,300 e-mails — and I draw from it a finding your logs were carrying without anyone having the time to establish it.
The gap measured, and it governs everything else: 12,200 of your 18,600 calls concern 15 topics, always the same ones — opening hours and days, school and canteen registration, identity card and passport, civil status, planning, bulky waste, cemetery, electoral rolls, municipal hall, social assistance from the social action centre, citizenship census, parking, weekly market, grants to associations, meals on wheels. Two calls out of three already have a written answer somewhere in your own files.
What that costs today, across the three items you already measure:
· At the Monday morning peak, waiting accounts for 60 % of a call's duration3 minutes out of the 5 the call takes from end to end.
· A routine enquiry accounts for 36 % of the time a request consumes at the switchboard3 minutes 36 out of 10.
· Booking an appointment with the right department: 48 %, that is 4 minutes 48.
· And the figure that holds the other three together: 4,650 calls out of 18,600 ring out unanswered — 25 %. That is not a reception failure: it is three officers for 18,600 calls, one of them part-time.
What I propose, and it is not an intention — it is already written: I drafted the 15 answers overnight, each backed by its source — resolution, decree, internal rules or website page — with its date in plain sight. 6 of the 15 answers your officers read today no longer match the organisation in force: two departments have changed reporting lines, one has moved premises, and three rates have been revised in council since. The 6 updated versions are written, with the replaced line and the line replacing it side by side.
What that shifts once the answers are settled: waiting drops to 10 %, the routine enquiry to 8 %, appointment booking to 9 %30 seconds, 48 seconds, 54 seconds.
The next step, and it takes twenty minutes: you read the 3 topics that weigh the most, I bring them to you tomorrow morning. The head of reception settles the version, and it is in service the same eveningthat approval is what commits the town's word, and it is the only act I leave to you. naubourg-switchboard_18600-calls-15-topics.pdf4,650 calls lost, 15 topics, 6 answers to revise
⛓ Sourced · 12 months of telephone system logs, service-user e-mails, resolutions and the town's website, 15 answers drafted
There are three of us on the front desk, one of them part-time. I do not have a whole day to spend on standard answers. Where do I start?
With three topics, not fifteen — and I have ranked them by the number of calls they take off you, not alphabetically.
What the count of your twelve months says:
· Opening hours, days and departmental contact details: 3,140 calls a yearthe switchboard's first topic, and the one whose answer never varies.
· School and canteen registration: 2,460 calls, packed into six weeks of the new school year, and it is the topic whose answer is out of date: the rate was revised in council and the front-desk notice still carries the previous one.
· Identity card and passport: 1,980 calls, of which most call for an answer from the town and an appointment, since you hold the enrolment equipment.
Those three topics make up 7,580 of the 12,200 recurring calls: three answers out of fifteen, six calls out of ten. The other 12 can wait until the end of the month without changing anything to your workload.
What I bring you within those twenty minutes, and what nobody has the time to do: the three updated answers, already written, with the governing resolution or page alongside and its date, and the number of calls each would have handled in the past year alone — 3,140, 2,460 and 1,980. You approve, or you correct with a single word. I save you the writing and the cross-checking: the decision takes twenty minutes instead of half a day, and it remains the town's — which is what makes it defensible before a service user.
The next step I propose: that after those three, I flag on my own any answer that a resolution, an organisational decree or a new rate amends, and that I write the updated version the same night. Your 15 answers will stop ageing in silenceand all it will cost you is one review, when the decision has been taken. fifteen-standard-answers_sourced-and-dated.pdf15 answers written, 6 updated, each with its source
⛓ Sourced · call count by topic over 12 months, register of resolutions, organisational decrees, 15 answers drafted
All of that assumes you hear what residents tell us on the phone. Where do their requests go?
They go nowhere. I run on local inference on a machine belonging to the authority, and nothing a service user tells me leaves your network.
Local inference means the model computes on your own machine: the service user's voice and the content of their request cross no external network to be processed. If the town would rather not host a machine, the other route is an isolated resource hosted in France, dedicated to Naubourgno pooling with another authority, which is the condition for the continuity of your service.
What that changes, point by point:
· Service users' requests train no model, neither ours nor a third party's.
· I work read-only on your files and calendars, and the technical account I read through has no right to writethat is sturdier than a promise, because it can be checked with one command.
· Encryption in transit and at rest, role-based accessrights follow the job: the switchboard does not open social assistance files, the planning officer does not open civil status. 7 roles for your departments, and the log shows 0 out-of-role access since go-live.
· Hosting in France, under French law, architecture designed to reduce exposure to extraterritorial legislation, location alone not being enough to guarantee immunity.
· Full log: which call, which topic, which answer given and from which source.
And one duty I honour from the first second, because it protects you as much as it informs the caller: I announce that I am a digital assistant of the town of Naubourg, not an officer. This is not an option you could switch off: the European regulation on artificial intelligence requires that anyone interacting with an AI system be informed. The caller can ask for an officer at any time — I transfer the call, and if the front desk is closed I take their number and drop the dated call-back into the department's queue.
The figure that sums all this up: 0 service-user data out of the town's network across the 18,600 calls of the year, and processing in the EU targeted.
What I propose: that I keep up to date the file your data protection officer and your council will ask for — hosting, data processed, retention periods for recordings, who accesses what. It is requested once a year and takes two days to reconstitute; the first version is already written and waiting for you. technical-framework_where-service-user-calls-live.pdfLocal inference, read-only, processing in the EU targeted
✎ Framework · deployment architecture, technical account rights, role matrix, first version of the register file
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 uses of AI for public phone reception

Each use corresponds to an agent we deploy. All of them work in support, under the supervision of a public officer.

Included in your agent The 2 capabilities essential to this promise are included, at no extra cost.

Switchboard & overflow

Picking up every call, absorbing peaks and overflow, transferring complex requests to an officer.

Appointment booking

Qualifying the request, scheduling the slot with the right department and sending the list of documents and the reminders.

Controls and safeguards These 5 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 Status, safe closure and audit trail Sources, access rights and handling of questions with no answer Offer accessibility, easy read, multichannel access and genuine human recourse Measure resolution, response time, non-take-up, drop-offs and satisfaction
Other needs our agents cover Each card says where the matching agent stands: available, on quote, or still being architected.

Accessibility and inclusion

To produce a plain-language version, prepare an easy-read transcript to the FALC method, translate or voice your content, this agent can be paired with the Accessibility and inclusion agent. None of these capabilities is included in what this offer covers as standard.

On quote View the agent page

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.

Does your need fall outside this?

In 15 minutes we identify the agent that will give your staff the most time back — without oversizing the project.

Book the free audit Build your agent
The gain

How much time can a switchboard win back?

By handling the first level and absorbing the peaks, an authority can aim for a sharp reduction in lost calls and waiting time — time given back to front-desk officers for the people present and the complex requests.

A caller's waiting time at peak hours
Today · done by hand
Immediate pick-up
Everyday information (opening hours, procedures)
Today · done by hand
Automatic
Booking an appointment with the right department
Today · done by hand
Near-instant
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

A switchboard voice agent (reception, information, routing, appointments), 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.

This agent is priced with you, not online. We are adjusting its scope at the moment, and online subscription stays closed while we do. Tell us what you need: we will come back to you with a price. Request a quote
Our commitment

Four guarantees that matter to a public service

The public's data never leaves the administrationLocal inference or an isolated resource hosted in France; no call data entrusted to a foreign third party.
Data in France, under French lawExposure of the public's data, hosted and minimised in France to the Cloud Act and FISA 702 is reduced by design; location alone does not guarantee immunity.
The public officer keeps the leadThe caller is told they are speaking to an AI; complex requests are transferred to a human officer.
Human oversight & traceabilitySwitchboard & overflow: monitoring, updates and call logging, compliant with the requirements of the AI Act.
Frequently asked questions

Your questions, our answers

Is the caller really speaking to a synthesised voice?
Yes, and they are told at the start of the call, as the AI Act requires. Complex or sensitive requests are transferred to a human officer: the AI agent assists reception, it never substitutes for the public officer on the subjects that need one.
Can we keep the existing number?
Yes: the agent connects to your telephony without changing your numbers. We adapt the integration to your phone system and your departments, without imposing a migration.
Does it work outside opening hours?
Yes, it can provide a first level of reception 24/7 — opening hours, procedures, contact details — and schedule call-backs and appointments for when the departments reopen, which strengthens continuity and equal access to public service.
Is the public's data protected?
Yes. Contact details and the content of requests are personal data: our agents run locally within the authority 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. GDPR & AI Act: governed deployment.
Can the agent really route to the right department?
Yes. From your organisation chart and your knowledge base, it qualifies the request and directs it to the competent department — civil registry, planning, social — or transfers directly to an officer. Every call is recorded for follow-up.
Does the agent state that it is an artificial intelligence?
Yes, from the very first interaction, 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 other party’s language, and they can ask for a human at any time.
How long does it take to deploy the agent?
A few weeks as a rule, after a free 15-minute audit that identifies the most useful use case, then a phase of design, integration with your telephony and testing before going live and training your front-desk officers.
Which tools can users use to reach the agent?
The ones they already have. The agent answers on WhatsApp Business, the website chat, email and VoIP telephony: users have no account to create and no application to install. This is a lever for access to the service — WhatsApp and the telephone 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. Sending the chosen slot and the list of documents to bring by text message, which saves a wasted trip 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, email or VoIP telephony. 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.
Let's talk

Let us estimate the potential for your switchboard

15 minutes to identify the most useful use case — hosted in France, supervised, with no commitment.