AI Systems Inventory for EU AI Act Compliance
You can’t comply with the EU AI Act if you don’t know what AI systems you’re using. Sounds obvious, but most teams don’t have a clear answer, especially when it comes to third-party tools, model APIs, and internal experiments. We will help you figure it out. We’ll work with your team to map out what counts as AI under the regulation, which ones fall under limited-risk or high-risk categories, tag the systems that matter, and give you a structure to keep it all in check going forward. No bureaucracy, no bloated frameworks — just a clear, usable overview.
We'll help you figure out what counts as an AI system, which ones actually matter for the EU AI Act, and how to keep track of it all without turning your architecture into a spreadsheet.
Schedule a call with usWhat's included in your AI inventory
Most teams are using more AI than they think — and some of it is already regulated. This assessment gets you from "we think we’re fine" to "we actually know."
- We go through your stack and workflows to find all internal and external AI systems
- We figure out what qualifies as AI under the EU AI Act, and assign risk levels to each system
- You get a structured, maintainable inventory — not a compliance artifact, but a real operational tool
This is where most of our clients start, because without an inventory, there’s no smart way forward.
What this helps you unlock
Once you know what AI systems you're working with and how risky they are under the regulation everything else gets easier:
- You'll know which systems require documentation, oversight, or transparency features
- You'll see where a AI literacy training, is needed — especially for teams building or integrating high- and limited-risk systems
- You'll be able to talk about compliance with confidence, not guesswork
Even if your team is "just using ChatGPT at work," it still counts, and it still needs to be inventoried. Here's why that matters under the EU AI Act.
This is how we help companies comply with the EU AI Act:
These blog posts and videos cover everything from understanding the law to applying it to ChatGPT use, AI literacy programs, and risk assessments.
Consultants
Paul Larsen
Head of AI at mkdev. Data scientist, PhD in mathematics and Rhodes Scholar with over 10 years of experience in both enterprise and startups. My hobbies are masters swimming, reading and walking our dogs.
and how we might help you
How it works
As there are no identical teams, projects and problems, so there are no identical solutions for all occasions.
We will find an approach that resolves your problem in a way that suits you best, and we will develop an ideal solution, given the needs of your business, your team and your product.
There is no unified solution, but in all cases, large and small, we will:
- get acquainted with your team, company and projects that you develop;
- carefully study the problem with which you encountered;
- prepare for you a solution and justify the need for certain actions, including in terms of costs.
Pricing
We develop solutions to various difficulties and scales, big and small. To achieve the best results, we apply an hourly rate for consultations. However, if we understand that to solve the problem a couple of hours is not enough, we will offer you a unique solution, carefully considering the needs of your business.