Sovereign generative AI

Your sovereign AI,
deployed on your terms

No dependency on OpenAI or Microsoft. You keep control of your data, EU hosting, GDPR and AI Act compliance. Proven by a three-week pilot, designed for public bodies and institutions.

GPU architecture · RAG · Mistral · OpenWebUI

Calculate your ROI See our references
413
Active users in production
200+
Concurrent users validated
100%
Sovereign EU hosting
The problem

Public AI is a data leak

Every prompt sent to a consumer AI is data leaving your security perimeter. Your staff are already doing it. The only question: within which framework?

Complete confidentiality

Your internal documents never pass through third-party servers. Interface on your side, AI in a European datacentre, a private encrypted tunnel between the two, with daily wiping of the processing servers.

GDPR & AI Act compliance

An architecture designed around the regulation: data in the EU, structural minimisation, DPO documentation provided, pseudonymisation available. You start compliant instead of catching up with compliance later.

Technical independence

100% open source components: no proprietary licence, no price rise imposed on you, complete reversibility. Your AI is your property, not a subscription.

Advanced RAG pipeline

From your documents to the sourced answer

Extraction (PDF, Office, OCR), vectorisation, re-ranking and inference: all of it locally, with citations of the passages used. Plus NotebookLM-style generations: summaries, FAQs, briefings, quizzes, mind maps and audio podcasts, 100% sovereign.

Your files all formats Extraction PDF·Office·OCR Chunking into chunks Vectors vector store Re-ranking BGE reranker LLM local inference Answer sourced + cited
Structural frugality

AI during working hours

The main GPU is rented at 08:00 and handed back at 18:00, wiped. Nights and weekends: off. Renewable energy, natural cooling in Nordic datacentres.

Energy avoided vs 24/7~70%

Multi-format

PDF, Office, OCR scans, source code. Every document becomes searchable.

Fixed cost

Shared infrastructure: 50 or 500 users, same invoice. No per-user licence.

Live monitoring

Prometheus & Grafana included, automatic alerts, daily backups.

Clients

Already in production in public bodies

No slideware: teams that use our sovereign AI every day, and talk about it.

City of Annemasse Annemasse Agglo
Territoire Innovant award · AI

City of Annemasse & Annemasse Agglo: "MIA"

In production for more than 12 months

Sovereign AI built on open source components (Mistral Small 3.2), hosted in Europe.

🏆 Awarded the "Territoire Innovant" label, AI category, by Les Interconnectés (June 2026, Paris)

340
active users
17,880
messages processed
1
single GPU

Official figures from June 2026: adoption keeps growing.

"With MIA, we are showing that a town can master its digital strategy without depending on outside players."
Gabriel Doublet, Mayor of Annemasse
annemasse.fr ↗ · one-year feedback (LinkedIn) ↗
Read the detailed case study
Calvados County Council
County-wide scale

Calvados County Council

In production for more than 6 months

The same sovereign foundation deployed across a whole county (~700,000 residents): an AI assistant connected to the county SSO, European hosting, open source models.

73
active users
12,940
messages processed
1
single GPU
Proof of scalability

From 90,000 to 700,000 residents: same architecture, same fixed cost, same sovereignty.

Simulator

What does it cost, what does it return?

The infrastructure is billed by the hour and only runs from Monday to Friday. Every assumption is editable: the calculation is instant.

Scenarios:
hours
€ / hour
Basis: average gross salary in French local government, €2,777/month (INSEE, 2023) + employer contributions
Advanced assumptions · infra · risk · compliance
€ / month 24/7
h → h
€ / month
€ / user / month
€
Conservative assumption: the French average is €3.59m (IBM Cost of a Data Breach, 2025)
% % reduced
How is this risk calculated?
€ / year
Legal advice or external audit avoided thanks to the compliance pack provided (DPO documentation, DPIA). In the high-risk category the AI Act obligations (conformity assessment, enhanced DPIA) are far heavier: × 3. Amount editable.
Operating hours / month—
Infrastructure adjusted to working hours—
Monthly cost of the sovereign solution—
Cost per user per month—
Equivalent in market licences—
Annual saving vs licences—
Value of time saved / year—
Break-even vs the market—
Data risk avoided / year—
Compliance avoided (DPIA, AI Act)—
Overall ROI (time + risk + compliance)—
🌱 Energy avoided (nights & weekends)—
🌱 CO₂ avoided vs a 24/7 service—

Method: shared GPU capacity, billed pro rata to working hours (reference: €1,400/month for 720 h round the clock; the reference price is the actual cost billed, not an on/off spot price) · risk = annual loss expectancy (cost × probability) · carbon: server ~1.2 kW including load, average European mix 300 g CO₂/kWh; our European hosting runs on renewable energy. Conservative assumptions, all adjustable.

Free tool · prioritisation

Where do you start?

Map the automation potential of your departments. Set the intensity department by department and get a prioritisation matrix, quick wins vs long projects, plus the top 5 automations to launch, with an estimate of the hours freed up.

  • Identify your quick wins department by department
  • Prioritise impact vs implementation effort
  • Move straight on to costing the ROI
Start the free audit
QUICK WINS PROJECTS Effort → Impact → 1 2 3 4 5
Preview; the real matrix fills in from your answers.
Technology stack

Auditable open source components

No proprietary dependency. Every component is replaceable and auditable.

The stack, deployed on your side

Every layer is open source and replaceable. The whole thing runs inside your perimeter, in the European Union, and nothing leaves for a third-party cloud.

YOUR SOVEREIGN INFRASTRUCTURE · EUROPEAN UNION Users · SSO sign-in Collaborative AI portal: OpenWebUI secure chat · knowledge bases · sharing by group RAG + Re-ranking multi-format extraction · embeddings · BGE reranker vLLM · inference engine open source LLM: Mistral Small 24B Sovereign GPU: EU datacentre · renewable energy Across the whole stack: observability (Grafana · Prometheus) · OpenAI-compatible API · daily backups
OpenAI-compatible endpoints: integrate with your business applications without ever leaving the perimeter.

Open source LLM

Mistral Small 3.2 24B Instruct in production, on a single GPU: sized to the actual need, no pointless gigantism.

vLLM Engine

High-performance inference, 200+ concurrent users validated in production.

Collaborative AI portal

Secure chat, knowledge bases, notes, sharing by group, SSO sign-in.

RAG + Re-ranking

Multi-format extraction, embeddings, BGE reranker: sourced and relevant answers.

Observability

Grafana & Prometheus, email alerting, daily backups, automatic failover.

API first

OpenAI-compatible endpoints to plug into your existing business applications.

Built for critical sectors

Designed for organisations that cannot afford a data leak.

Public bodies Banking & finance Healthcare
Defence
Education
Legal
The PixelRaise team

Technical credibility forged by expertise

PixelRaise does not sell licences: we deploy and maintain your own sovereign infrastructure. Based in the Annemasse area, with teams in France and Switzerland.

⚡ Method: three weeks to decide on facts

From scoping to the go/no-go milestone, with no commitment to roll out.

W1

Scoping & installation

Audit of the use cases, SSO connection, sovereign infrastructure in place, first document bases imported.

W2–3

Pilot with 20 users: go/no-go milestone

Real usage, adoption and time saved measured. You decide on the numbers.

Go

Roll-out

Opened to all users, up and running in 30 minutes, in-house champions trained.

Test sovereign AI on your own documents

A 30-minute demonstration on a sample of your real documents, then a three-week pilot (go/no-go milestone included). An expert replies within 48h.

Book a demo · 30 min

Pick a slot directly, or fill in the form alongside.

info@pixelraise.com
251 Route des Marais, 74250 Fillinges, France
Resources

Understanding sovereign AI

Clear guides for your technical teams and your decision-makers: from model technology to regulatory compliance.

Digital sovereignty

The three pillars of sovereign AI: data location, compliance, control over the model.

Sovereign LLM models

Open source vs open weight, an agnostic stack, model comparison (Mistral Small 4…).

AI & GDPR

Legal bases, DPIA, and the right to erasure inside a RAG system.

EU AI Act

Risk levels, obligations for deployers, 2026 deadlines.

AI API anonymisation

Protecting sensitive data even with ChatGPT, Claude or Gemini. Coming soon.

ROI calculation method

How we put a figure on the data risk avoided (annual loss expectancy).