AIMS Agent: your AI Management System, documented
AIMS is short for AI Management System — the record-keeping structure the EU AI Act expects and ISO/IEC 42001 describes. AIMS Agent builds yours: the documentation Regulation (EU) 2024/1689 asks for, produced from what you know about your own systems, in eight European languages today and all 24 in time.
What it is for
The AI Act does not ask for a statement that you comply. It asks for records: what the system is, which risk category it falls in, how that was decided, what governs it, and what happens when it changes.
AIMS Agent builds that structure from a description of your systems, and keeps it current as they change.
What it produces
Risk classification
Which category a system falls in under the Act — and, more usefully, the reasoning and the Articles behind it, so the answer can be checked rather than trusted.
Technical documentation
The records the regulation expects a provider or deployer to hold, drawn from your answers rather than from a template you have to fill in yourself.
Eight languages, then 24
English, German, French, Dutch, Italian, Spanish, Polish and Portuguese today. All 24 official languages are the goal, because every version of the regulation is equally authentic and your regulator reads one of them.
Which role are you?
The Act does not treat everyone the same. What you owe depends on your relationship to the system — whether you built it, deploy it, import it, distribute it, or represent a provider based outside the Union. Articles 22 to 25 set out those duties separately.
AIMS Agent covers all five roles: provider, deployer, importer, distributor and authorised representative. It establishes which apply to you from your own answers, rather than asking you to know in advance — and the answer changes what it produces. An EU-established provider and a non-EU provider without a representative do not receive the same set of documents, because they do not owe the same things.
How it works
You describe your AI systems. AIMS Agent asks the questions the regulation makes relevant to them — not every question, only the ones your answers make necessary — and produces the documentation from what you tell it, with the Articles it relied on attached.
What that looks like
Say a company is building a tool that screens job applications. It describes what the system does. AIMS Agent matches that against Annex III and returns high-risk under Annex III, point 4(a) — “AI systems intended to be used for the recruitment or selection of natural persons, in particular to place targeted job advertisements, to analyse and filter job applications, and to evaluate candidates” — with the reasoning and the citation attached.
From there it produces the technical documentation Article 11 requires: what the system does, how it was tested, how it is monitored. Drawn up before the system is placed on the market, and kept up to date.
The system is built around what it has to be able to account for, so the record does not only say what your compliance position is — it says how it got there, and when it changed. That property is difficult to add later, which is why it is in the foundation rather than bolted on.
What it will not do
It will not tell you that you comply. No software can, and any that claims to is selling you a risk it cannot carry. What it does is produce the documentation and the reasoning, so that the people who are accountable can review a real argument instead of assembling one from scratch.
Where it runs
AIMS Agent is a hosted application at aimsagent.ai. This page explains it; that is where the work happens.
Not yet generally available. If the AI Act applies to you, we would rather talk early than have you find out late.