EU AI Act readiness in six weeks
Illustrative example. Figures are representative targets, not a specific client claim.
A fast-moving scale-up had shipped AI into customer-facing features before anyone had mapped the compliance picture. There was no inventory of what used AI, no risk classification, and no documentation, a board-level exposure under the EU AI Act, and a growing worry that fixing it would mean freezing the roadmap.
The real risk wasn't the technology
The systems themselves were fine. The exposure was that nobody could say, on paper, which features used AI, what data they touched, or what would happen if a model got something wrong, exactly the questions a regulator or enterprise customer asks first.
Left alone, that gap tends to surface at the worst possible moment: during due diligence, a security review, or an incident, when there's no time to assemble the answers.
How we'd approach it
We'd inventory every system that uses AI, classify each by risk under the EU AI Act, and write lightweight governance policy that fits how the team already works rather than bolting on heavy process.
The output is documentation and simple approval workflows that make the AI defensible without slowing delivery, so the product team keeps shipping while the board gets the assurance it needs.
A scale-up deploying AI in customer-facing products had no inventory, classification or documentation, a board-level risk.
We inventoried every AI system, classified risk under the EU AI Act, wrote lightweight governance policy and set up documentation and approval workflows.
A defensible, fully documented AI governance framework the board signed off on, without slowing the product team down.
- Start with an inventory: you can't govern AI you haven't mapped.
- Classify by risk so effort goes where the Act actually requires it.
- Aim for lightweight, living documentation that survives contact with a fast roadmap.
Not sure which applies to you?
Book a free assessment and we'll map the highest-ROI automation opportunities for your business, honestly, including when it's not worth starting yet.
Book a free AI assessment