EU AI Act readiness: risk classification, documentation, transparency and human oversight
We put all your AI systems into one inventory, help determine their EU AI Act risk category and build the technical side: documentation, logs, user transparency and human oversight.
What hurts
You sell in the EU or work with European partners, and they already ask about the EU AI Act. You do not know which of your systems fall under it or what to do about it.
Lawyers explain the regulation text but cannot tell you how to log model decisions or where to get a description of training data. The technical side stays empty.
AI tools appear in departments on their own. Nobody has a full list of them, who owns each one or what data goes in.
What we build
- AI system inventory: in-house builds, vendor services and department initiatives
- Preliminary EU AI Act risk classification of each system together with your counsel
- Determining your role for each system: provider or deployer
- Technical documentation: purpose, data, model, limitations and tests
- Logging of system decisions and actions for later review
- Transparency: notices that a person is talking to AI and explanations for users
- Human oversight and escalation procedures for disputed cases
- A gap remediation plan with priorities and deadlines
What you get
- An AI system register with risk categories and owners
- A technical documentation pack for each higher-risk system
- Logging and transparency mechanisms set up in the product
- A gap report and a work plan
- Training for the responsible staff
How we measure the result
- Share of AI systems in the register with a named owner
- Number of gaps closed from the plan
- Documentation and logs in place for each higher-risk system
- Time to answer a partner or regulator request
Systems and stack
Industries
Questions
Do you give legal advice on the EU AI Act?
No. We are an engineering team and prepare the technical side: the register, documentation, logs and controls. The final legal classification of systems and any legal opinion come from your counsel or an external adviser. We work alongside them and give them the facts about your systems.
How much does EU AI Act readiness cost?
From $7,000 over 4–8 weeks. The price depends on the number of AI systems, their risk category and how much documentation already exists. One higher-risk system needs more work than ten minimal-risk chatbots.
Does the EU AI Act apply to companies outside the EU?
The regulation covers systems placed on the EU market or used in the EU, wherever the developer is based. If you sell in the EU or your system output is used there, it is worth checking. Whether it applies to you is finally decided by your counsel.
What do we need to provide?
A list of known AI systems, access to their documentation and code, owners in each department and your counsel or DPO. We sign an NDA and a GDPR DPA before access. We operate under ISO/IEC 27001:2022.
How does your ISO 27001 help?
Many EU AI Act expectations on logging, access control and risk management overlap with ISO 27001. We went through that process and know how to present evidence. But our certificate covers our company, not yours.
What happens after readiness work?
Documentation has to match how the system works today. We can update the register and documents under an SLA whenever systems change, or hand the process to your owners with templates.
Related services
More in this section
Describe the task. We reply within one business day, the first call is with the founder.
Describe the task. The first call is with the founder.
We reply within one business day. 30 minutes to tell you whether it is worth doing and what it costs.