01 · KNOWLEDGE
Information exists, but cannot be used reliably
Documents are distributed, outdated or contradictory. Employees keep searching or asking the same experts.
AI readiness · data & processes · personally delivered
Your AI will not fail because of the model. It will fail because of data and processes.
I help mid-sized companies organise knowledge, data and workflows so they can build AI solutions that work reliably in day-to-day operations.
10+ years in complex automotive projects
Diagnosis and technical delivery from one source
Mid-sized companies · project organisations · knowledge work
The real problem
A chatbot is quick to build. A reliable knowledge base is not.
Information is scattered across emails, PDFs, network drives, Confluence, Excel and the minds of individual employees. Nobody knows for certain which version is valid, who is responsible or whether the AI may access it.
Automating chaos only produces automated chaos.
That is why I do not begin with a tool. I begin with processes, data and ownership—and turn them into a solution that is traceable, maintainable and commercially sound.
02 · PROCESSES
Manual handovers consume capacity every day
Data is copied, status reports are compiled by hand and tasks are transferred between systems.
03 · OWNERSHIP
Nobody owns the complete system
Business teams, IT and service providers each see only one part. A working end-to-end solution is missing.
The core offer
The AI Readiness Sprint
In ten working days, a vague AI idea becomes a robust basis for decision-making and implementation—including a focused technical pilot for the most important use case.
- Process and bottleneck analysis
- Data sources and access
- Data quality assessment
- Prioritised AI use case
- Technical proof of concept
- Business case
- Target architecture
- 90-day implementation plan
The process
Clarity first. Technology second.
STEP 01
Focus
We select a process with genuine commercial leverage—not the most spectacular AI trend.
STEP 02
Data
I assess sources, versions, quality, access, data protection and functional ownership.
STEP 03
Pilot
A focused prototype using real data shows what works and where the limits are.
STEP 04
Decision
You receive costs, benefits, risks and a concrete implementation plan—not a generic presentation.
Client case · anonymised
In this AI chatbot project, AI was not the bottleneck.
The chatbot itself was straightforward to implement. The decisive work came first: locating knowledge, assessing versions, clarifying access and structuring content so that answers become reliable and traceable.
Starting pointScattered documents, uncertain validity and unclear ownership.
Decisive leverA curated knowledge base with sources, owners and an update process.
OutcomeAn AI system that does not merely generate text, but works with dependable company knowledge.
Typical starting points
Where the sprint creates particular value.
Company knowledge
Internal assistants and chatbots with traceable sources instead of unsupported AI answers.
Documents and back office
Read, validate, classify and transfer incoming documents into existing systems.
Project information
Bring together status, decisions and tasks from Jira, Confluence, email and spreadsheets.
10+years working on systems and projects where failure was not an option.
Why me
Engineering thinking meets hands-on delivery.
I worked on steering and driver-assistance systems at Volkswagen and led international projects. That environment teaches you that a strong presentation is no substitute for a functioning system.
Today I combine this process and project experience with automation, AI and software development. I also operate my own digital products and businesses. That is why I consider not only the technology, but also effort, adoption, operations and commercial viability.
Engineering instead of consulting: I diagnose, structure, build and stay involved until the system works in practice.
Frequently asked questions
Before our first conversation.
Is the sprint already a complete AI project?
No. It reduces the risk before a larger investment. You receive a focused pilot and all the foundations needed to make an informed decision about implementation, budget and partners.
Does our data need to be prepared already?
No. Assessing and structuring the available data is part of the sprint. The relevant people and data sources only need to be accessible in principle.
What happens after the ten working days?
You can continue internally, work with your existing providers or ask me to deliver the next stage. The sprint does not commit you to a follow-up engagement.
Can sensitive data be processed locally?
Depending on the use case, the solution can be designed for local operation, a controlled cloud environment or a hybrid setup. Data protection, access controls and technical boundaries are explicitly addressed.
Do you still offer interim project management?
Yes—particularly when a complex digital or automation project needs operational leadership. The AI Readiness Sprint is simply the clearly defined entry point for new AI and automation initiatives.
Personal introductory call
In 30 minutes, we will determine whether your challenge is a good fit for the sprint.
You will receive an honest initial assessment of the potential leverage, your data situation and the right next step. If the sprint is not the right choice, I will tell you just as clearly.
Directly with me. No obligation. A concrete first assessment.