FAQ
Frequently asked questions.
What is EA Brain?
EA Brain is our flagship product, an AI enterprise architect. It turns conversation into documentation and diagrams, maintains a living knowledge graph of the company and tracks compliance with regulations such as DORA, NIS2, GDPR and the AI Act. It also serves as the knowledge base that agents draw on.
Who does the AI Act apply to, and since when?
Most companies use AI rather than build models, so only a smaller part of the obligations applies to them. The employee AI literacy obligation has applied since February 2025, and the remaining user obligations take effect in August 2026. We handle compliance together with the deployment, not as a separate project.
How does an engagement start?
We always start with a workshop to find out what your company actually needs. Then comes a small pilot on one process in real operation, with a success criterion agreed in advance. Only after that do we expand to further departments, add the governance layer and train your people.
Do we have to commit to a large project right away?
No. We proceed in three stages, from a company assistant through company knowledge to agentic work. The stages follow one another and each has its own deliverable, so you can stop at any point and lose nothing.
How do you handle the security of AI agents?
An agent with access to company systems is a new type of user with permissions that need to be clearly defined and guarded. We offer an AI Security Assessment as well as DevSecOps for AI development. Agents are tested and approved before they get access to systems.
Where do you operate?
Primarily in Slovakia, with projects across the EU. We work in Slovak and English.
Which AI models do you use?
We are not tied to a single vendor and we don't sell licences. We choose the model based on the task and the sensitivity of the data. All models go through one gateway, so access, logging, limits and costs stay in one place regardless of which model is used.
How long does a deployment take?
A company assistant can be deployed within weeks, including a usage and cost overview from day one. The pilot has a success criterion agreed in advance, so you quickly know whether it makes sense to continue. Later stages depend on scope and data readiness.
What happens to our data?
Before deployment we agree on rules for which data may enter a model, and verifiable records are kept of how it was used. GDPR compliance and an audit trail are part of the solution, so you can always show what happened with your data.
How much does it cost?
The price depends on scope, which is why we start with a workshop and a small pilot with a clear criterion rather than a big project bought blind. Thanks to the FinOps layer, operating costs are visible by department, with budgets and caps, so nothing surprises you.
Can we run it ourselves later?
Yes, that is part of the plan. During the operations phase we transfer competence to your team, and we help grow key users in every unit who move their own tasks to AI and teach their colleagues.
How do we fulfil the employee AI literacy obligation?
With short practical training and clear rules on what is and isn't allowed. We train leadership, key users and the whole company, and completion is verifiable, exactly as the AI Act requires.
Can we see references?
References on the website are anonymised because we don't name our clients without their consent, and we will protect you the same way. We can say much more in a personal meeting.
Will AI replace our employees?
Agents take over routine work, but they hand important steps to a person for approval. The goal is to move teams towards more valuable work, which is why training and adoption are part of every deployment, not a replacement of people.