AI consulting that goes all the way to running systems.
We find out together where AI really helps your business, and then I build it. Strategy, knowledge systems and automation for the midmarket, from Dusseldorf across NRW.
How I build.
Privacy by default, not by request.
Particularly in Europe, AI implementation requires an acute awareness of data boundaries. I architect systems that keep sensitive information on-premise or firmly within contracted boundaries.
Do not automate broken processes.
Putting AI on top of a messy workflow just creates a faster mess. Strategy means looking at the operational logic first, fixing the foundation, and automating second.
The prototype is the strategy.
You learn more by building a functional minimum viable product than by looking at five slide decks. Execution uncovers constraints that strategy alone cannot see.
Eliminate friction, keep humanity.
Technology should absorb repetitive administrative friction so humans can focus on deep work, strategy, and empathy. Not the other way around.
How to start.
A clear ladder, not a flat menu: initial call, audit, project or pilot.
Initial call → Audit → Project/Pilot. Audit is the paid diagnosis step, Pilot the implementation after. Not either-or, but a path.
AI Potential Audit
Für wen
You want to know where AI gives your team time back and pays off as ROI before you invest big.
Vorgehen
I map your processes, check feasibility and data boundaries, prioritize by effort and value.
Ergebnis
A clear map of which use cases are worth it and what the next step is.
Audit Kompakt: 1.500 € netto
1 Tag, konkrete Landkarte
Audit Tiefen: 2.400 € netto
2 Tage, mit ROI und Roadmap
Both audits are credited on commissioning a project.
Handoff
As handoff you get an AI strategy plan with concrete integration options. You can continue yourself, hire another provider, or me. Easiest with me because I already know the context and credit the audit fee.
Knowledge System / RAG
Für wen
Your knowledge lives in many tools and few heads. You want it searchable without sending sensitive data to the cloud.
Vorgehen
I build a local or GDPR-compliant retrieval system over your documents. Answers with sources, without data leaving your network.
Ergebnis
Your team asks in plain language and gets cited answers from your own materials.
Agentic Workflow / Automation
Für wen
Repetitive manual work eats time, tools exist but are not connected.
Vorgehen
I connect your systems into automated workflows that run independently.
Ergebnis
Less manual work, fewer errors, a workflow your team understands and operates.
How my projects run.
Discovery and scoping
We look at your processes, data and goals. Result: clear problem and scope.
Strategy plan
You get a prioritized map of what is worth it. Result: plan and proposal.
MVP build
I build the smallest useful version. Result: something that runs.
Pilot and QA
We test in real daily life and sharpen. Result: robust system.
Rollout and handoff
Your team takes over, with enablement. Result: no permanent dependency.
Your team runs it themselves. No permanent dependency.
What sets me apart from standard tools and big consultancies.
| Kriterium | Standard-Tool / große Beratung | Meine Arbeitsweise |
|---|---|---|
| Privacy | US servers, often no DPA | German or EU servers, or local. GDPR documented |
| Usage | Isolated tool, individual employees | Integrated into your processes |
| Accountability | No clear process ownership | Concrete use case with measurable goal |
| Implementation | Slides, then another provider | Strategy and build from one hand |
| Independence | Selling own product lines | Vendor-neutral, you get a free strategy plan |
| Success measurement | Unclear | KPIs defined before project start |
GDPR is architecture to me, not a label.
Local or EU servers
AI can run entirely on your hardware or on German and EU servers, without data flowing to a US cloud.
Data boundaries clarified
In every project we clarify data boundaries, access roles and deletion rules upfront.
DPA and processing register
A data processing agreement and a simple processing register are included.
Local when needed
If public cloud is not an option, we build locally.
What if only one person does everything?
A fair question for a solo provider. My answer is structure over dependency.
I document every system so your team can operate it themselves, and hand it over consciously at the end. You get no blackbox consultant, but a system that runs without me. For ongoing operations we agree clear maintenance windows if needed.
One proof.
No real client cases yet. The proof comes from my own systems, which I run daily and share openly.
View systemsPilot projects 2026.
I am currently taking a small number of pilot clients. We develop your first AI project together, at reduced rates. You get a running system and full attention from someone who consults and builds. In return I may use the case as a reference, anonymous on request.
Example topics
- →Internal knowledge system (RAG)
- →Automating a manual process
- →A custom agent
- →Or your own topic, we clarify in a call.
Nothing fits?
Book a free initial call. We find out together whether and how I can help.
Start a conversation