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

AI agent for the municipal police (administrative support)

An assistant that prepares the orders, lays out the reports and keeps the administrative incident log — the officer and the authority keep every decision. A strictly administrative and documentary scope: no operational, biometric or surveillance use. Hosted in France, on a resource isolated for each authority.

Hosted in France Strictly administrative support GDPR & AI Act: governed deployment Human oversight

Updated on

Deployed in a few weeks
Municipal police assistant · hosted in France
Prepare a draft order banning parking on rue Pasteur on Tuesday, for a house move.
Draft order ready. Citation of articles L.2213-1 and L.2213-2 of the local government code, grounds (house move with a van), area (rue Pasteur, even-numbered side, from no. 8 to no. 14), hours 7am – 6pm, signage at the applicant's expense.
Still to complete: the order number — for the competent authority to approve and sign.
⛓ Source · your order templates + the mayor's policing powers
Lay out this morning's report from my notes: fly-tipping observed on chemin des Vignes.
Report laid out. Observations dated and time-stamped, precise location, description of the facts, photographs referenced, a documentary reminder of the possible administrative follow-up. No legal classification is settled on your behalf.
The administrative incident log has been updated at the same time.
✎ Action · report ready for review — the officer checks and signs it
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 decidesAdministrative support, never operational
✦ In brief

For the municipal police, a Blue Lemon Agent agent works exclusively in administrative supportdraft municipal orders, reports, incident logs, minutes, replies to residents — and keeps a watch on the mayor's policing powers. No operational, biometric or surveillance use: that scope is excluded by design and by the AI Act. It runs on local inference or is hosted in France on a resource dedicated to and isolated for each authority, architecture designed to reduce exposure to extraterritorial legislation, location alone not being enough to guarantee immunity. Every document is reviewed and signed by the officer or the competent authority; the time won back goes to being present on the ground. Live within a few weeks. Your public-sector staff write to it from Microsoft Teams, Slack or their email, and residents 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 municipal police forces — and why they hesitate

Orders, reports, incident logs, registers: the administrative side of the job keeps growing, at the expense of being present on the streets. And the subject is sensitive: AI belongs here only in documentary support, never on the ground.

! The issue

Municipal police officers spend a growing share of their time at a desk — draft orders, reports, incident logs, minutes — at the expense of being on the ground, which is the heart of the job. Yet most consumer AI tools would amount to entrusting observations, identities and residents' circumstances to a third party, often hosted outside Europe and subject to the Cloud Act.

Our answer

For the municipal police, AI is only acceptable if it is sovereign, confidential and strictly confined to administrative support. Local inference or an isolated resource hosted in France, systematic human oversight, no operational, biometric or surveillance use — excluded by design and by the AI Act. The time saved on drafting goes back to being present on the streets; every act stays decided and signed by the officer or the competent authority.

The decisive point

Protecting the data and the terms of use: sovereignty & compliance

Reports and incident logs contain sensitive data about residents. Here is how the architecture of our agents protects it — and how the scope stays strictly administrative.

Local inference

The agent can run on a machine belonging to the authority: 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 — residents' data: processing and access within the European Union targeted by the architecture.

Reduced extraterritorial exposure

Exposure of residents' data to the Cloud Act and FISA 702 is reduced by design; location alone does not guarantee immunity.

One isolated resource per authority

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

The agent is strictly in administrative support; no operational, biometric or surveillance use; 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

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.

The public body in this demonstration

Fictional public body

Municipality of Roche-l'Étang — 21,600 inhabitants, municipal police station (fictional municipality)

Sector
Municipal police of a town of 21,600 inhabitants — 47 km of municipal roads, a weekly market, 38 public-highway events declared in the year, 4 school groups
Headcount
14 staff at the station, including the 11 municipal police officers and the head of service in the pilot, 2 public-highway wardens and an administrative officer; the urban CCTV control room stays outside the scope, and the agent has no access to it
Public served
6,400 enquiries from residents a year — 4,100 calls to the station, 1,400 emails, 620 counter visits, 280 website forms; lost property, parking, noise, animals, permits and formalities lead the list
Order of magnitude
640 draft orders prepared a year, 1,860 reports and administrative incident-log entries, 52 weekly activity reports, 86 legal texts and rulings tracked by the watch
Tools already in place
The station's incident-log software, the register of orders, order templates settled by the chief executive of the municipality, the town's road plan, electronic signature folder, Microsoft Teams and email — the agent plugs into them read-only, nothing is replaced or migrated
Who decides what
The mayor holds the police powers and signs the orders — CGCT, art. L2212-2, L2213-1 and L2213-2; the municipal police officer records breaches by procès-verbal and reports to the judicial police officer — Internal Security Code art. L511-1, Code of Criminal Procedure art. 21-2; the head of service settles the templates. The agent prepares, writes up and dates; it characterises no fact
Room for improvement
1,148 calls out of 4,100 ring unanswered — 28.0 %; a resident's email waits 9 days on average; activity reports and administrative follow-up are 100 % manual, 3 h 30 a week and 182 hours over the year; 96 draft orders out of 640 had to be redone before signature — 15.0 %

At Roche-l'Étang, eleven municipal police officers cover 47 kilometres of roads, a weekly market and 38 declared events a year. The desk takes a growing share of the job: 640 draft orders, 1,860 reports and incident-log entries, a weekly activity report, and 6,400 enquiries from residents, most of them on a dozen topics that keep coming back. The exchanges below cover one year, from the survey of the flow to the review presented to the municipal council. The agent is hosted in France on a resource isolated per municipality, reads the incident log and the register of orders without ever writing operational content to them, and its scope is strictly administrative and documentary: it prepares orders that the mayor signs, writes up reports from what the officers saw, serves residents' routine enquiries — and it has no access to CCTV feeds, performs no biometric recognition and characterises no fact.

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.

Municipal police assistant · hosted in France
I have reread your 6,400 enquiries for the year and gone back over your 640 draft orders against your register and your road plan — the review that 47 kilometres of roads never leave time for.
What the count says: 10 topics cover 4,260 enquiries out of 6,400, that is 66.6 % of the flow. Lost property, obstructive parking and impounding, neighbourhood noise, stray animals, permits to occupy public space, temporary drinks licences, removal-day orders, burning and dumping, holiday property watch, formalities and documents required. The 10 reference answers are written, sourced on your municipal orders in force and on the legal texts, and dated — they are waiting for you.
What the three items weigh, and you can check each on your own logs:
· A draft order takes 34 minutes end to end, of which 60 % is re-entering the template, the citations and the perimeter — 20 minutes 24.
· A report or incident-log entry takes 14 minutes, of which 40 % is writing up — 5 minutes 36.
· Activity reports and administrative follow-up are 100 % manual work: 3 h 30 a week for the head of service, that is 182 hours over the year.
And the survey that pleases nobody, starting with me: 96 of your 640 draft orders were redone before signature — 15.0 %. I will tell you straight away that most of those redos are my doing, and the next exchange tells you exactly which and why.
The proposal, and it takes half an hour of your time: I have written you four control rules for draft orders and run them over your last 24 months — 1,280 orders. The head of service settles the wording, and they run on the March batch. station-flow_6400-enquiries-10-topics.pdf4,260 enquiries across 10 topics, the three items costed, 456 hours identified
⛓ Sourced · 12 months of enquiry log, register of orders, municipal road plan, electronic signature folder
Those 96 redos — is it us writing the request badly, or you writing the order badly?
It is me, and on one precise point. 71 of the 96 redos — 74.0 % — turn on the PERIMETER. I was carrying over from the template the phrase « rue Pasteur », which is the applicant's phrase, whereas article L2213-2 of the French general code of local authorities makes parking a measure taken « by reasoned order »: an order banning parking along a whole street for a two-hour removal is an order that gets challenged, and your officers were correcting it by hand, every time.
What I did about it: since 14 April, the perimeter is requested and written in street numbers, with the length in metres and the number of bays concerned, taken from your road plan. Across the 214 drafts produced since, 7 were redone — 3.3 % — and none on that ground.
And here are the four control rules, run over your 1,280 orders of the last 24 months. For each one, the number of flags, the share confirmed on examination by the head of service, and what it lets through:
· Rule A — perimeter with no street numbers or length: 186 flags, 179 confirmed — 96.2 %.
· Rule B — temporary order with no end date, or whose duration exceeds that of the stated reason: 142 flags, 128 confirmed — 90.1 %.
· Rule C — citation of an article that is not the basis of the stated reason: 98 flags, 74 confirmed — 75.5 %. It is the least reliable of the four, and I am giving it to you as such: in 24 cases the broad citation was a deliberate choice by the service.
And here is what those three let through, because you need to know it before you sign: 61 orders have no date of posting or publication in the register — so their entry into force is not established, and none of the three sees them. I wrote rule D — order signed with no publicity date entered in the register within 48 hours — and ran it over the same 24 months: 61 flags, 58 confirmed — 95.1 %.
The four rules together cover 402 of the period's 431 fragile orders — 93.3 %, that is 201 a year. Bringing them into service stays with the station: a rule only takes effect once the head of service has validated it. four-control-rules-for-orders_tested-over-24-months.pdf4 rules tested, 402 fragile orders covered out of 431, the 71 perimeter redos and their fix
⛓ Sourced · CGCT art. L2213-2 · 24 months of the register of orders, 1,280 orders, signature-folder redo log
And the watch on the mayor's police powers? We stopped it last year for lack of time.
It is running again, on five declared sources, and it tracks 86 texts and rulings a year on your scope: Légifrance for the general code of local authorities and the Internal Security Code, the Official Journal, prefectoral orders in the register of administrative acts, published administrative case law on the mayor's police powers, and your own municipal council resolutions.
What I publish against myself, because that fifth source was missing: out of the year's 86 texts, my watch missed 5 — 5.8 %. And the explanation is particularly awkward for me: all five are resolutions of YOUR municipal council. I was watching national law and not watching the town. Two of the five amended the weekly market regulations, which three of your order templates cite.
What I did about it: the internal source has been connected since 2 May, and I caught up the previous 24 months — 34 resolutions recovered, 6 of which touch a template in service. Those six templates are marked « to be reviewed » in the signature folder; I did not modify them, the head of service and the chief executive settle the templates.
The useful marker: the delay between a text or a resolution being published and it reaching your templates has gone from 21 days to under 24 hours, and it no longer depends on someone having the time. What I propose: a two-page watch note every Monday — what has been published, what it touches here, which template is concerned — and you decide whether you want the administrative case-law review, which adds a quarter of an hour of reading. police-powers-watch_5-sources-86-texts-and-the-5-missed.pdfThe 5 sources, the 5 resolutions missed and their cause, the 24-month catch-up
⛓ Sourced · five declared watch sources, 24 months caught up, the station's register of order templates
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 support of the municipal police

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

Included in your agent The 5 capabilities essential to this promise are included, at no extra cost.
From 790 € incl. VAT / month

Draft municipal orders

Traffic, parking and event orders prepared from your templates, ready to sign.

Reports & incident logs

Laying out the reports and the administrative incident log from the information entered.

Replies to residents

Answer everyday questions (lost property, permits, procedures) and direct people, 24/7.

Minutes & registers

Activity minutes and administrative registers kept up to date from the information supplied.

Watch on policing powers

A targeted watch on the mayor's policing powers and the applicable procedures, sourced and dated.

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 Work from a versioned corpus with citations and the law as it stood on a given date Preserve confidentiality, compartmentalisation and access logging Manage deadlines, versions, evidence and human validation Flag uncertainties and reserve advice, decision and signature for the lawyer
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
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 municipal police station win back?

By automating the preparation of orders, reports and minutes, a town can aim for a clear reduction in desk time — reinvested in being present on the streets.

Drafting a draft order
Today · done by hand
Prepared by the agent, to approve
Laying out a report or an incident log entry
Today · done by hand
Near-instant
A week's minutes and administrative tracking
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

An administrative support agent (draft orders, reports, incident logs, replies to residents), 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

8,450 € incl. VAT setup
then 790 € 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

1,260 € 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

13,286 € 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)
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 municipal police force

Residents' data never leaves the councilLocal inference or an isolated resource hosted in France; no data entrusted to a foreign third party.
Data in France, under French lawResidents' data: minimisation and location in France, architecture designed to reduce exposure to extraterritorial legislation, location alone not being enough to guarantee immunity.
The municipal police officer keeps the decisionThe agent prepares orders and reports that can be checked; no operational, biometric or surveillance use.
Human oversight & traceabilityDraft municipal orders: every change to the service is logged, in line with the requirements of the AI Act.
Frequently asked questions

Your questions, our answers

Is the agent used on the ground, or to identify people?
No: strictly administrative and documentary. No operational, biometric or surveillance use — excluded by design and by the AI Act.
Where is the data processed?
In France, on a resource isolated for each authority, with logging and role-based access. GDPR: governed deployment.
Does the agent have an operational or surveillance use?
No: its scope is strictly administrative support (drafting, tracking); no surveillance, recognition or operational decision function.
Who approves the orders produced?
The public officer: the drafts are reviewed and signed by 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 change software?
No. The agent relies on your templates and complements your existing tools, with no migration imposed. We adapt the integration to your environment.
Which tools can residents use to reach the agent?
The ones they already have. The agent answers on WhatsApp Business, the website chat and email: a resident has no account to create and no application to install. The scope stays administrative and non-urgent — lost property, permits, procedures: the agent gives out the emergency numbers and never steps in for the officers on a situation that calls for them. 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.