Services
Four ways to put AI to work for your organisation. Every engagement starts with a free call and ends with a fixed-price proposal — you know the cost before we start.
AI Act readiness, EU hosting, governance
Sovereign AI & Compliance
For organisations that must keep AI inside the EU and inside the rules — banks, insurers, the public sector, health. We design AI systems your regulator, your DPO and your board can sign off on. Everything we ship is EU-hosted and GDPR-first, and our work is built on Luxembourg public open data — Lux Memorial is a registered reuse on data.public.lu.
AI Act Readiness Audit
We inventory your AI uses, classify each one against the EU AI Act risk tiers, and deliver a gap report with a remediation roadmap. Fixed price.
Sovereign AI Architecture
EU-hosted or self-hosted LLM systems: model selection, hosting design, data flows that never leave the EU — and DPIA input your DPO can work with. Fixed price.
AI Governance Retainer
Ongoing advisory as AI Act obligations phase in — prohibited practices are already in force, GPAI rules apply since August 2025, and high-risk obligations arrive in August 2026 and 2027 — plus vendor reviews and a compliance watch. Monthly.
FAQ
Does the AI Act already apply to us?
Partly — and more of it each year. The Act entered into force in August 2024 and applies in phases: prohibited practices and AI-literacy duties since February 2025, obligations for general-purpose AI models since August 2025, and most high-risk obligations from August 2026, with some extending into 2027. Which obligations apply to you depends on how your AI uses are classified — that is exactly what the readiness audit determines.
Can we use US AI providers and stay GDPR-compliant?
Sometimes — it depends on your data flows, your data-processing agreements and the current state of EU–US transfer arrangements, and it needs case-by-case review. A sovereign design removes the question: when models run on EU infrastructure — or your own — there is no transatlantic transfer to justify.
Do you replace our lawyers?
No. We do the technical and architectural side: the classification groundwork, the system design and the documentation your counsel and DPO build on. Legal interpretation stays with your lawyers — we work with them, not instead of them.
Not sure which one you need?
That's what the free call is for. Describe the problem — we'll tell you what fits, and what it costs.
Strategy, audit, roadmap
AI consulting
Before you spend money building anything, know what AI can realistically do for your organisation. We look at your processes, your data and your constraints, and give you a plan you can act on — with us or without us.
AI strategy
Where AI genuinely helps your business — and where it doesn't. Prioritised by impact and feasibility, not by hype.
AI audit
A review of your existing tools, data and processes: what's automatable today, what needs groundwork first, and what the risks are (GDPR included).
Roadmap
A concrete, sequenced plan: which project first, what it needs, what it should cost, and how to measure whether it worked.
FAQ
Do I need an AI strategy before building anything?
Not always. If you have one clear, well-scoped problem, we can go straight to a fixed-price proposal for building it. Consulting is for organisations that know AI matters but don't yet know where to start or in what order.
What does an AI audit look at?
Your processes, the data you actually have (not the data you wish you had), the tools your team already uses, and your regulatory constraints such as GDPR. The output is a written assessment of what is automatable now and what needs groundwork first.
Is the advice tied to specific vendors or models?
No. We recommend what fits your case — commercial or open-source models, EU-hosted by default. If the honest answer is that you don't need AI for a problem, we say so.
Not sure which one you need?
That's what the free call is for. Describe the problem — we'll tell you what fits, and what it costs.
Apps, chatbots, automation, MCP servers
Custom AI systems
When the plan is clear, we build the system: scoped by the founder, built by our international team of specialists assembled for your project, and handed over working. EU-hosted and GDPR-compliant by default.
AI applications & chatbots
Customer-facing assistants and internal tools that answer in your users' language — like the multilingual concierge we built for a hospitality venue in a Luxembourg thermal resort.
Business automation
AI-driven document processing and workflow automation — the repetitive reading, sorting and extracting your team shouldn't do by hand.
MCP servers
Connect AI assistants to your real business data — CRMs, databases, internal tools — with access control and audit logging.
FAQ
How long does a custom AI system take to build?
Typical projects deliver in weeks, not months. The fixed-price proposal you receive after the first call includes a delivery date, and that date is part of the agreement.
Where does the system run and who owns it?
On EU infrastructure by default, and you own what we build. We document what is processed and why, so your own compliance or legal team can sign it off.
What is an MCP server and do I need one?
MCP (Model Context Protocol) is an open standard that lets AI assistants securely use your business systems — read your data, act in your tools. You need one when you want AI that works with your data rather than a generic chatbot.
What happens after delivery?
You get a working, documented system and a handover. Ongoing support and monitoring are available as a separate, clearly-priced agreement — never an obligation.
Not sure which one you need?
That's what the free call is for. Describe the problem — we'll tell you what fits, and what it costs.
RAG, search, pipelines
Your data, AI-ready
AI is only as good as the data it can reach. We make your documents and databases usable by AI — so answers come from your actual contracts, policies and manuals, with sources cited, instead of a model's guess.
RAG systems
Retrieval-augmented generation: the AI answers from your documents and cites where each answer came from.
Search
Semantic search over your documents and knowledge bases — find things by meaning, not just keywords.
Data pipelines
The unglamorous foundation: extracting, cleaning and structuring your data so AI systems stay accurate as your business changes.
FAQ
What is RAG, in plain words?
Retrieval-augmented generation. Instead of answering from memory, the AI first retrieves the relevant passages from your own documents and answers from those — with citations. It's the difference between an assistant that read your contracts and one that guesses.
Our documents are a mess. Is that a problem?
It's the normal starting point. Part of the work is exactly this: extracting, cleaning and structuring what you have. We tell you in the proposal how much groundwork your data needs — before you commit.
Does our data leave the EU?
No — EU hosting is our default, and we document the full processing chain so you can verify it. If a specific tool would require transfers outside the EU, we flag it and propose an alternative.
Not sure which one you need?
That's what the free call is for. Describe the problem — we'll tell you what fits, and what it costs.