US

Working with US companies

An EU-based AI architect for US companies working with European data.

I design AI agents, automation and business tools for US teams whose products or operations involve personal data of people in the European Union. European constraints are part of the scoping from day one: where the data lives, who can access it, and what the model is allowed to do.

Why European data changes the design.

The GDPR can apply to a company established outside the EU when it offers goods or services to people in the EU, or monitors their behavior there (GDPR, Article 3(2)). An AI feature that reads customer messages, documents or records is then a processing operation like any other: it needs a purpose, a legal basis, a known chain of processors and a retention rule.

The EU AI Act can also apply to providers and deployers established outside the Union when the output of an AI system is used in the EU (AI Act, Article 2). Whether and how it applies depends on the system and its use; that assessment belongs to your legal counsel. My role is to make the technical choices visible and documented, so they can be assessed.

European constraints, handled as engineering.

  • Architecture choices documented before development

    Data flows, hosting location, model provider, access rights and logs are described up front, so your privacy and security teams can review them before anything is built.

  • European hosting or local models when the data requires it

    When data must stay in the EU or inside your own environment, the system can rely on a provider hosted in Europe or on a model deployed on your infrastructure, rather than sending everything to a remote API.

  • AI agents with explicit boundaries

    What the agent can read, which actions it can take, what requires a human confirmation and when it must stop are defined and tested, not left to the model.

  • Evidence rather than promises

    Evaluation on your own cases, references traced back to their sources and limits stated plainly. The goal is a system you can explain to a customer or an auditor.

Remote collaboration, one point of contact.

  • Remote, in English

    Workshops by video call, written documentation and demonstrations on your own cases. Deliverables can be written in English.

  • Overlapping hours with the East Coast

    I work from France, on Central European Time: Paris afternoons overlap with US East Coast mornings.

  • One point of contact

    You work with me from the first audit to go-live and training, without hand-offs between intermediaries.

What we can build together.

Each service has its own page: situations, deliverables, process and questions.

  • Sovereign AI

    The right starting point if you need to decide on use cases and architecture.

  • On-premise LLMs

    The right starting point if your hosting constraint is already clear.

  • Document automation

    The right starting point if a document task ties up your teams.

  • AI agents & API integrations

    The right starting point if your need involves actions across several tools.

  • AI training for teams

    The right starting point if your teams already use AI without shared guidelines, or are hesitant to start.

  • Custom business applications

    The right starting point if a spreadsheet, ill-fitting software or repeated data entry is slowing down an important process.

  • AI assistant for your documents

    The right starting point if your teams spend a long time looking for information that already exists in your documents.

Case studies with European data constraints.

Your questions.

Can you guarantee GDPR or AI Act compliance?

No. Compliance is assessed by your organization and its legal counsel, for a given system and a given use. I design and document the technical side (data flows, hosting, access, logs, model choices) so that this assessment rests on facts.

Do you work with US-based model providers?

Yes, when your constraints allow it. The choice between a US API, a provider hosted in Europe and a local model depends on the data involved, your contracts and the performance you need. The trade-offs are written down during scoping.

Do we need to be based in Europe to work together?

No. The work happens remotely. What matters is whether your product or operations involve data of people in the European Union, and how much control you need over that data.

How do we start?

Describe your product or process, the data involved and what you want the AI to do. A first conversation helps choose between a scoping audit, a build or training.

07

Next step

Let’s discuss your project.

A few lines on your context, the data involved and the outcome you are looking for are enough to start.

Discuss your project