FR Book a call
Home Solutions Full AI audit Software: MIA Custom AI API Training & Conferences About Resources FAQ Use case library Enterprise AI blog Book a call FR
100% on-premise

Your company.
Your data.
Your AI.

The enterprise AI assistant deployed on your own servers, enriched by your internal documents. Flat rate, zero tokens, zero Cloud Act.

Discover the solution Contact us
Access to Mia from your local network
mia.artemia.ai
AssistantsQuestions rapidesDocuments
Thomas V.
Summarise the remote-work agreement in 5 points
Mia is reading your documents…
Mia
• Up to 3 days a week, subject to manager approval
• Available after 6 months of seniority
• Flat monthly allowance of €72.60
remote-work-agreement.pdfhr-note-2026.docx
Processed on your server · nothing leaves
Start a conversation with Mia.
4.5/5 on Trustpilot
Digital sovereignty Local deployment, outside the Cloud Act
Multi-model LLMs updated on a regular basis
Unlimited tokens Flat fee, unlimited usage
Trusted by
Previsia
Yvelines, Le Département
Ville de Châtillon
Solvéo Énergie
Comptastar
CGA Midi-Pyrénées
Kleos
Fidsud
Lot-et-Garonne, Le Département
GT Company
Cotelib
V and B
Our approach

From audit to automation,
AI deployed on your premises.

  • Public sector
  • Private companies
  • SMEs & mid-caps
  1. 01 · Understand AI audit

    Know where AI will pay off.

    A clear picture of your uses, your data and your teams, to know where generative AI will really pay off.

    • A map of AI uses, shadow AI included
    • Risk review: confidentiality, GDPR, EU AI Act
    • Prioritised use cases and a roadmap
    Opportunity portfolio Prioritising
    Use caseValueFeasibilityComplianceDecision
    1. Legal research Wave 2
    2. Marketing video avatar Ruled out
    3. Meeting summaries Wave 1
    4. CV screening Ruled out
    5. Citizen enquiries Wave 1
    6. Incoming mail triage Wave 2
    Wave 1 · 2Wave 2 · 2Ruled out · 2
    Discover the audit
  2. 02 · Deploy MIA

    Generative AI for every team.

    The generative AI software that puts your internal documents to work — without ever letting them leave your infrastructure.

    • RAG over your internal bases, answers that cite sources
    • Collaborative assistants per team or department
    • Deployed on-site in 24h
    mia.your-company.com
    AssistantsQuick questionsDocuments
    Thomas V.

    Write the job posting for a maintenance technician.

    MIA is reading your documents…
    Mia

    Done — the posting follows your hiring guidelines:

    W maintenance-technician-posting.docxWord · 2 pages Open
    Maintenance technician
    Your role
    Your profile
    What we offer
    hiring-guidelines.pdf maintenance-job-sheet.docx
    Processed on your server · not one byte left the network
    Thomas V.

    Compare our Q1 and Q2 expenses, line by line.

    MIA is reading your documents…
    Mia

    Here is the variance by line, from the general ledger:

    LineQ1Q2Change
    Energy€42,300€38,900−8%
    Subcontracting€61,800€70,200+14%
    Travel€12,400€9,700−22%
    general-ledger-2026.fec
    Processed on your server · not one byte left the network
    Thomas V.

    Which public tenders in industrial maintenance this week?

    MIA is searching the web…
    Mia

    Three notices published since Monday:

    1. Waste-processing authority — sorting lines, €420k1
    2. Regional hospital — multi-trade maintenance, 3 years2
    3. Seaport — crane servicing, lot 21
    1 boamp.fr 2 ted.europa.eu
    Secure web search · no tracking, no profiling
    Thomas V.

    Generate the September production report as a PDF.

    MIA is reading your documents…
    Mia

    The report is ready, charts included:

    Production report — September production-report-september.pdf PDF · 6 pages · 412 KB
    mes-export-september.csv targets-2026.xlsx
    Processed on your server · not one byte left the network
    Thomas V.

    Write a script that merges this month’s CSV exports.

    MIA is reading your documents…
    Mia

    Here is a Python script, ready to run:

    merge_exports.pyPython
    import pandas as pdfrom pathlib import Path files = sorted(Path("exports").glob("*.csv"))df = pd.concat(pd.read_csv(f) for f in files)df.to_excel("exports-september.xlsx", index=False)
    internal-dev-guide.md
    Processed on your server · not one byte left the network
    Start a conversation with Mia.
    Discover MIA
  3. 03 · Automate Custom API

    AI wired into your tools and agents.

    The server, the gateway and the keys: an OpenAI- and Anthropic-compatible AI running inside your walls.

    • GPU server and gateway delivered and installed on-site
    • OpenAI- and Anthropic-compatible: one URL, one key
    • IDEs, Claude Code, n8n, local agents: everything plugs in
    Automated workflow Running
    Your network
    Trigger E-mail received Quote request — Dupont Ltd
    API Artemia Mia-Agent Reading the message…
    Actions CRM record updated Reply drafted Follow-up scheduled
    24 s · 4,820 tokens
    Trigger PDF invoice F-2041 · Supplier Build+
    API Artemia Mia-Agent Extracting amounts…
    Actions Amounts extracted Ledger entry created Approval requested
    39 s · 7,350 tokens
    Trigger IT ticket “The VPN is down”
    API Artemia Mia-Agent Diagnosing…
    Actions Ticket tagged: Network Fix suggested Technician notified
    44 s · 9,610 tokens
    Discover the API
04 · Why on-premise

What changes with sovereign AI

Both keep the same functional promise. The difference shows once AI is in production: in your data, your costs and your independence.

Read the on-premise AI guide
Criterion
Generic AI solutions
Model size
Artemia Right-sized models, matched to your use cases*
Cloud AILarger models, trained to do everything
Data confidentiality
Artemia Total — on-premise deployment
Cloud AIData sent to the cloud
Exposure to US law
Artemia None — outside the Cloud Act
Cloud AISubject to the Cloud Act
Pricing model
Artemia Flat fee, unlimited usage
Cloud AIPer user or per token
Cost predictability
Artemia Total — fixed costs
Cloud AIVaries with usage
Advertising in the interface
Artemia None — you are the customer, not the product
Cloud AIBeing introduced on consumer plans
AI architecture
Artemia Multi-model, vendor agnostic
Cloud AISingle proprietary models
Heavy production usage
Artemia Optimised and cost-effective at scale
Cloud AIHigh costs at scale

* In favour of ethical, frugal AI: deploying a model able to handle 96% of all your professional tasks (internal benchmark across more than 25 models and 360 business use cases, against the legacy model).

AI audit · MIA · Custom API

Book a call with an expert

Find where AI will pay off, deploy it inside your walls, then automate your processes: in one call, we review your challenges and show you where to start.

  • A conversation about your business challenges
  • A demo tailored to your context
  • A roadmap, with no commitment
Book a call

A question first? Read the FAQ · info@artemia.ai