Rules for AI, before the auditor demands them.
A governance framework for AI systems that fits the EU AI Act and data protection and works in daily practice.
Once AI works with real data, rules are needed
Once AI agents or models work with real data and real decisions, informal use is no longer enough. The EU AI Act classifies applications by risk, data protection requires a legal basis for processing, and internally someone has to be accountable when a system makes a mistake.
Many companies either have no rules at all, which creates risk, or a policy nobody reads because it has nothing to do with daily practice. We build a framework of a few clear rules: which use cases are allowed, who approves them, which data may be used, and how to respond if something goes wrong.
The result isn't a document for the drawer, but a process that runs automatically with every new use case: risk classification, approval, documentation.
Risk classification, approvals, and documentation in minutes
Risk classification
Classification of your AI use cases according to the EU AI Act's risk categories.
Approval process
A clear process for who reviews and approves new AI use cases before they go live.
Data protection framework
Rules for handling personal and confidential data in prompts and training data.
Responsibilities
Naming who is functionally and technically accountable for which AI system.
Documentation template
A template that documents each use case in a few minutes, instead of tying up a compliance department.
Training for responsible parties
Short, practical briefing for the people who grant approvals or operate systems.
From the use case inventory to an anchored approval process
- 01
Take stock of use cases
We capture which AI systems are already in use, officially or unofficially.
- 02
Classify risk
We assign each use case to the EU AI Act's risk categories.
- 03
Design the framework
We design the approval process, data protection rules, and responsibilities together with leadership and compliance.
- 04
Anchor it in the company
We train the responsible parties and embed the process where new use cases arise.
Leadership, compliance, and advisory boards
| Situation | Our approach |
|---|---|
| AI usage is growing without rules | A lean framework creates control without slowing down innovation |
| Advisory board or investor asks about governance | A solid answer with documentation instead of an ad-hoc explanation |
| Several portfolio companies with different levels of AI maturity | A shared minimum standard every company can implement |
Our approach: A lean framework creates control without slowing down innovation
Our approach: A solid answer with documentation instead of an ad-hoc explanation
Our approach: A shared minimum standard every company can implement
No legal advice, only the operational framework
We're not a law firm and don't give binding legal advice on the EU AI Act. We build the operational framework and work with your legal department or external counsel as needed, rather than replacing their role.
Frequently asked questions
| Ausgangslage | Umsetzung | Ergebnis | |
|---|---|---|---|
| EdTech company, PE-financed | PE-financed EdTech company, technical picture unclear before the investment. | Tech DD, greenfield architecture, AI agents in production, AI-assisted engineering organization. | The company runs AI in day-to-day operations, not just in pilot projects. |
| EdTech company, customer service | Customer service answered recurring questions manually from scattered knowledge. | AI agent with access to the knowledge base, with clear escalation to humans. | Recurring inquiries run productively through the agent, complex cases stay with the team. |
| Manufacturing company, family-run | Business teams tested AI tools without a shared overview or approval rules. | An AI inventory, documented risk reviews, an approval process and role-specific training. | Owners can review new uses and track changes to existing applications. |
Ausgangslage
PE-financed EdTech company, technical picture unclear before the investment.
Umsetzung
Tech DD, greenfield architecture, AI agents in production, AI-assisted engineering organization.
Ergebnis
The company runs AI in day-to-day operations, not just in pilot projects.
EdTech company, customer service
Ausgangslage
Customer service answered recurring questions manually from scattered knowledge.
Umsetzung
AI agent with access to the knowledge base, with clear escalation to humans.
Ergebnis
Recurring inquiries run productively through the agent, complex cases stay with the team.
Manufacturing company, family-run
Ausgangslage
Business teams tested AI tools without a shared overview or approval rules.
Umsetzung
An AI inventory, documented risk reviews, an approval process and role-specific training.
Ergebnis
Owners can review new uses and track changes to existing applications.
A first conversation takes 30 minutes.
We show what a lean governance framework could look like for your AI use cases.