+33 (0)1 87 66 00 65 · Monday to Friday, 9am–6pm Free audit (15 min)
● Public sector — Civil status registrar

AI agent for the civil status registrar

An assistant that prepares the records, adds the marginal notes and handles requests for copies — the registrar checks and signs every record. Hosted in France, on a resource dedicated to and isolated for each town, civil status data stays under French law. The AI agent assists, the registrar decides.

Hosted in France Civil status data protected GDPR & AI Act: governed deployment Human oversight

Updated on

Deployed in a few weeks
Civil status assistant · hosted in France
Prepare the draft marriage record for file no. 2026-114, ceremony on Saturday at 11am.
Draft record ready. Identities, addresses and witnesses taken from the file, publication of the banns verified (the ten-day period observed), no interview required on the basis of the file, in line with your templates.
Still to complete: the mention of the marriage contract if the notary's certificate is produced before the ceremony — for the registrar to check and sign.
⛓ Source · the marriage file + your record templates
A notary's office is asking for a full copy of a 1998 birth record.
Request checked. The notary establishes their standing: the full copy may be issued. Record located in the 1998 register, copy prepared with the marginal notes up to date, covering letter drafted.
Nothing goes out without your authorisation.
✎ Action · copy prepared — the registrar checks and signs
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 registrar decidesThe AI agent assists, it never signs alone
✦ In brief

In the civil status department, a Blue Lemon Agent agent assists the registrar with the repetitive tasks — draft birth, marriage and death records, marginal notes, full copies and extracts, replies to the public — and checks the consistency of the documents. It runs on local inference or is hosted in France on a resource dedicated to and isolated for each town: the registers and the civil status data are never exposed to a foreign service, architecture designed to reduce exposure to extraterritorial legislation, location alone not being enough to guarantee immunity. The registrar checks and signs every record; the time won back goes to welcoming families. A light deployment, available by direct award below the public procurement thresholds. 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
7
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 civil status departments — and why they hesitate

Records to draw up without error, marginal notes to add within the deadlines, requests for copies piling up: civil status demands absolute rigour, on the most identifying data there is — births, marriages, deaths, parentage.

! The issue

The registrar is caught between families and professionals (notaries, lawyers, public bodies) who expect records that are quick and beyond reproach, and a demanding formality where every error can have lasting legal consequences. Yet most consumer AI tools would amount to entrusting births, marriages, deaths and parentage to a third party, often hosted outside Europe and subject to the Cloud Act.

Our answer

For civil status, 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, signature reserved to the registrar: the time saved on preparing records is never paid for in lost confidentiality, nor in a break in the continuity of the service. The aim is not to replace the registrar, but to give them back time for families.

The decisive point

Protecting civil status data: sovereignty & compliance

A civil status department handles the most identifying data of the town's residents. Here is how the architecture of our agents protects it, town by town.

Local inference

The agent can run on a machine belonging to the town hall: no 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 civil status data: processing and access within the European Union targeted by the architecture.

Reduced extraterritorial exposure

Civil status data: architecture designed to reduce exposure to extraterritorial legislation, location alone not being enough to guarantee immunity.

One isolated resource per council

No pooling of data: an environment strictly dedicated to your town, guaranteeing the continuity of the 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 act signed 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, 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 Beaumont-la-Rivière — a council of 18,400 residents, with a maternity unit on its territory

Sector
Urban council — civil registration, cemeteries and elections service; one maternity unit on its territory; counter open 35 hours a week
Headcount
210 staff, 5 of them in the civil registration service: one service manager and four case officers; the mayor and three deputy mayors are registrars
Public served
18,400 residents — and far beyond: the maternity unit sees 1,240 children born a year, 890 of them to residents of other towns
Volume
1,736 records drawn up a year (1,240 births, 186 marriages, 214 deaths, 96 acknowledgements of parentage), 128 civil partnerships registered, 3,100 marginal annotations, 14,800 requests for copies and abstracts, 18,300 information enquiries
Tools in place
Population and civil registration software, registers digitised since 1905 — 214,000 records —, an online request form and email; the agent plugs into them read-only, nothing is replaced and nothing is migrated
Who decides
The mayor and the deputy mayors sign the records; the civil registration service manager signs off every draft before signature; nothing leaves the service without a sign-off
Room for improvement
1,460 marginal annotations waiting, entered in 21 days on average; a copy issued in 8 days; 38 % of calls unanswered; and 4,800 of the 214,000 digitised records are still hunted for by hand

In Beaumont-la-Rivière, five staff run the civil registration of a territory that produces far more records than its population would suggest: the maternity unit makes the town the place of birth of 890 children from other towns every year. Formality there is absolute, and that is exactly what a document-by-document match secures: a forename carried onto a record follows a person for life, and it now reaches the sign-off with the document that establishes it alongside. The agent runs on local inference on a machine at the town hall, reads the registers and the files read-only, and puts forward every draft with the document that establishes it and the place where that document says so. The exchanges that follow cover one year, from clearing the annotation backlog to the review presented to the chief executive's office.

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.

Civil status assistant · hosted in France
I indexed your digitised registers over the weekend — 214,000 records since 1905 — and two figures come out of it that your service was already carrying without ever having had the time to set them side by side.
Indexing means reading the content of every record so that it can be found by what it says — a name, a date, a parentage — and not by register number and year. Your registers stay exactly what they are: nothing is copied, nothing is moved, no record is altered.
First figure, the one that decides: 1,460 marginal annotations are waiting to be entered, 318 of them for more than six months. A marginal annotation is the entry made in the margin of a record to flag a later event — marriage, divorce, death — so that the record always states the current situation. An annotation still waiting shows up at the counter, the day a family asks for a copy — and the 318 oldest are already matched, drafted and ready to sign off.
Second figure: 4,800 of your 214,000 digitised records are still hunted for by hand — 2.2 %. They were digitised as plain images. I ran 1,200 of them through overnight as a trial: 1,140 are now found by what they say, and the other 3,600 follow at the same pace as soon as you tell me to.
What those three areas weigh today, and what they will weigh tomorrow:
· Preparing a draft record takes 60 % of the time a record needs15 minutes out of the 25 a record demands, from file to sign-off.
· Handling a request for a copy or an abstract: 40 %, that is 4 minutes out of 10.
· A week's marginal annotations: the whole of the time given to them, that is 5 hours.
What I do with those three, and it is already in place: I prepare them, every element accompanied by the document that establishes it and the place where that document says so, and your service keeps what it alone can do — check and sign off. Preparation drops to 10 %, 10 % and 20 %: 2 minutes 30 seconds, 1 minute, 1 hour a week.
And the backlog of 1,460 annotations, I bring you sorted rather than merely stated: the 318 oldest are matched to their record, drafted in your service's form and set out as a comparison list. At 45 sign-offs a day, they are behind you in seven working daystell me to start with them and Monday's list is ready tonight. register-status_214000-records-indexed.pdf214,000 records findable, 318 annotations ready for sign-off
⛓ Sourced · registers digitised since 1905, annotation notices received and awaiting entry, the service's log
We know about the annotation backlog. What we do not know is where to take hold of it without bringing the rest of the service to a halt.
By the 318 oldest, and without halting anything at all: the backlog clears in seven weeks at one hour of sign-off a day.
What I have done with the stock, by sorting it rather than stacking it:
· 1,010 marriage and divorce annotations — the simplest to match up: the notice received carries the same identities as the record it targets. All 1,010 are drafted and awaiting sign-off.
· 200 death annotations, drafted the same way.
· 250 annotations of acknowledgement of parentage, or change of surname or forenamethese touch parentage and identity: I present them apart, one at a time, the notice and the record side by side.
The arithmetic, which you can redo: 1,460 annotations prepared in batches, signed off at 45 a day — that is what your service manager sustained through the trial week. Seven weeks, and the stock is at zero, without an hour of extra help and without touching the current flow, which I prepare in parallel.
What reaches your sign-off reaches it finished: the match made, the notice on one side, the record on the other, the discrepancies made plain, the annotation already drafted in your service's form. The sign-off carries 45 annotations into the register in one signatureand it is what gives the annotation the same force as the record.
What I propose next: that the current flow be prepared on the day the notice arrives, rather than whenever there is time for it. The time to enter an annotation goes from 21 days to 4and a family asking for a copy receives a record that states today's situation. The setting is ready: it takes effect on tomorrow morning's first notice. annotation-backlog_1460-sorted.pdf318 more than six months old, seven working days of sign-off to clear them
⛓ Sourced · 1,460 annotation notices sorted by type and by age, the trial sign-off week, the log of turnaround times
All of this assumes you read our registers and our files. That is the most identifying data we hold. Where does it go?
It stays with you. I run on local inference on a machine at the town hall, and nothing I read leaves the council's network.
Local inference means the model computes on your own machine: the text of a birth record or of a marriage file crosses zero outside networks to be processed. If the council chooses not to administer a machine of its own, the other route is an isolated resource hosted in France, dedicated to Beaumont-la-Rivièreno pooling with any other council.
What that changes, point by point:
· Your registers train zero models, neither ours nor a third party's.
· I work read-only, and the technical account I read through is stripped of any right to write into the civil registration softwarewhich is sturdier than a promise, because it can be checked in the configuration.
· Encryption in transit and at rest, role-based accessrights follow the job: a front-desk officer opens copy requests, not acknowledgement-of-parentage files; a cemeteries officer opens concessions, not birth records. An access right is withdrawn on a word, and the withdrawal takes effect within the minute.
· Hosting in France, under French law, architecture designed to reduce exposure to extraterritorial legislation, location alone not being enough to guarantee immunity.
· A complete audit log: who asked for what, when, which record was consulted and what the system produced.
And what gives your records their legal force: they are drawn up and signed by a registrar who answers for them before the publicthat is what makes them authentic and binding on everyone. What I bring to each of those acts is the work already done: a draft record built with its document alongside, an annotation drafted and matched, a copy prepared with its annotations up to date.
And on correction, a figure the indexing turned up: 62 records in your registers carry a divergence with another record in the same holdings — a forename, a date, a parentage. All 62 files are built: the two records side by side, the chronology of the documents, the exact place of the divergence and the draft referral to the competent authority. The decision to correct belongs to that authority — and I hand it over ready, instead of leaving it to be assembled.
What I propose: that I keep up to date the sheet your data protection officer will ask for — data processed, retention periods, who accesses what, where it is hosted. On civil registration that is the first question you will be asked, and it is written before anyone asks it. technical-framework_where-your-registers-live.pdfLocal inference, read-only, processing in the EU targeted
✎ Framework · deployment architecture, rights of the technical account, 62 correction files built
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 in the civil status department

Each use corresponds to an agent we deploy. All work in support, subject to the registrar's approval.

Included in your agent The 7 capabilities essential to this promise are included, at no extra cost.
From 1,035 € incl. VAT / month

Draft records

Birth, marriage and death records prepared from your templates, to be checked and signed.

Marginal notes

Preparation of the notes (marriage, divorce, death) to be added once the registrar has checked them.

Copies & extracts

Handling requests for full copies and extracts, with a check on the requester's standing.

Checking documents

Verifying the consistency of the documents in a file (marriage, acknowledgement of parentage) and flagging discrepancies.

Answers to the public

Answer everyday questions (documents, timescales, civil status procedures) and direct people, 24/7 and across channels.

Appointment booking

Schedule civil status appointments (filing a marriage application, acknowledgement of parentage, civil naming ceremony).

Plain language & easy-read draft

Preparation of a plain-language version and a draft easy-read (FALC) transcription, subject to human validation.

Controls and safeguards These 7 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 Prepare without deciding: sourced rules, supporting documents, anomalies and explanation Guarantee the decision, signature, recourse and responsibility of the public official Record versions, access, criteria, actions and notifications Test for bias, false positives, fundamental rights and continuity of service
Other needs our agents cover Each card says where the matching agent stands: available, on quote, or still being architected.

Assisted drafting of official acts and letters

Orders, resolutions, notes, standard letters. Human approval.

On quote View the agent page
The gain

How much time can a civil status department win back?

By automating the preparation of records, notes and copies, a town can aim for a clear reduction in time spent on repetitive tasks — reinvested in welcoming families and checking the files.

Preparing a draft record
Today · done by hand
Prepared by the agent, to approve
Handling a request for a copy or an extract
Today · done by hand
Near-instant
Updating a week's marginal notes
Today · done by hand
Prepared by the agent, to approve
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 civil status agent (draft records, marginal notes, copies and extracts), installed and operated for you. Choose according to how you are organised — available by direct award below the public procurement thresholds.

Agility

Setup + controlled subscription

12,025 € incl. VAT setup
then 1,035 € 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
Order →
The simplest Serenity

All inclusive, no setup fee

1,705 € 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
Order →
100% Sovereign

On site, you own it

17,061 € incl. VAT setup
then 1,300 € 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)
Order →
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 civil status department

Civil status data never leaves the townLocal inference or an isolated resource hosted in France; no data entrusted to a foreign third party.
Data in France, under French lawExposure of civil status data to the Cloud Act and FISA 702 is reduced by design, that data being located and minimised in France; location alone does not guarantee immunity.
The registrar keeps the decisionThe agent prepares records, notes and copies that can be checked; no signature is automated.
Human oversight & traceabilityFull traceability for the draft records: monitoring, updates and logging compliant with the AI Act.
Frequently asked questions

Your questions, our answers

Does the agent sign the records?
No: it prepares; the registrar checks and signs. Legal responsibility remains entirely human.
Is it compatible with our population software?
Yes: secure integration with your population and civil status management software, defined at the design stage.
Where are the digitised registers kept?
In France, on a resource dedicated to and isolated for each town, under French law — architecture designed to reduce exposure to extraterritorial legislation, location alone not being enough to guarantee immunity.
Does the agent handle corrections to records?
It prepares the drafts and spots the errors to correct, but the correction (administrative or judicial) remains the decision of the competent authority.
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 an agent?
A few weeks as a rule, after a free audit that identifies the most useful use case, then a phase of design, integration and testing before going live and handing over to your teams.
Do we need a technical team in-house?
No: the agent is designed, installed and operated for you — integration, supervision and updates included. Your teams concentrate on their work, with guided familiarisation.
Do we have to run a procurement procedure?
Below the public procurement thresholds, the agent is available by direct award. We supply what your decision needs: a quote, GDPR and AI Act documentation, and the hosting arrangements in France.
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 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 — are passed on at actual cost, with no margin, outside the subscription.
Let's talk

Let us estimate the potential in your council

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