The supplier decided to start in small steps, with the option to stop after each one. That is exactly how the collaboration is built. Three stages, each with a clear result, the first step costs nothing.
Three stages, each with a defined result. You can stop after each stage, and you should be able to decide after each stage whether the next is worthwhile.
The sequence is deliberately chosen so that the first stage lies below the usual investment threshold. This lets you bypass the planning and approval apparatus that delays most projects by a year.
Stage 1, Orientation
What happens. A conversation of about an hour. We go through the starting situation, the occasion, the process behind it and the question of whether there is a lever at all. I also tell you if there is none.
What you have afterwards. A documented, clearly laid-out assessment. It names which use case is most likely to hold up, which risk class it falls into, which prerequisites are missing at your company and what a sensible next step would be.
Effort. One appointment plus preparation.
Investment. The first consultation is free of charge.
Stage 2, Data classification and pilot
This is the decisive stage. Here it becomes clear what would later block everything else.
What happens. A moderated appointment with management, the department, IT and, if needed, the works council. We classify your data types into four classes and determine which tools are permissible for which class.
A use case is then set up on existing hardware and tested with your real documents. No purchase, no server procurement, no connection to specialist systems.
What you have afterwards. A documented data rule on one page, with a named person for doubtful cases. Plus a reliable result on whether the use case holds up, with which model and with what effort. And a recommendation on whether the step into regular operation is worthwhile.
Effort. Half a day on site plus setup and evaluation.
Investment. On request.
Important. This stage is deliberately cut so that it fits into a current budget year and does not trigger an investment application.
Stage 3, Support of the rollout
What happens. If the pilot holds up, it moves into regular operation. Scope and duration depend on which expansion stage you are aiming for, that is, whether one area, several areas or a connection to ERP or product lifecycle management (PLM).
What is included. Selection and procurement of the hardware, building the environment, the authorisation concept, connecting the data sources, training the users and the documentation you need for evidence.
What I do not do. I do not operate the system permanently for you. The goal is that your IT can take it over. If this capacity is not available, we say so early and look for another way.
Effort. Project scope, depending on the expansion stage.
Investment. On request, depending on the expansion stage.
Training
Anyone who uses AI must, under the EU AI Act, ensure sufficient AI competence among the people working with it. What sufficient means depends on the use. Someone who has texts drafted needs different knowledge than someone who evaluates test protocols or processes customer data.
General online courses convey a basic understanding. For documentation, what counts is the reference to your company, not a certificate of attendance.
My training therefore starts from your actual use cases and from the data rule that arose in stage 2. It can be booked as a separate module or embedded in stage 3. More on this under Training.
Investment. On request, each including preparation and follow-up.
We discuss concrete terms and funding options in the first consultation.
Funding options
Consulting services are eligible for funding depending on the federal state and the project. Whether a programme comes into question for your case, I look at together with you, and I support you with the classification and the prerequisites. What fits concretely depends on the federal state, company size and project, which is why we clarify it in the first consultation before we begin.
Talk to me about it, and we check it before we start.
What you should bring
For a first conversation, a concrete occasion is enough. You need no strategy, no data preparation and no idea of which tool comes into question.
It helps if you can say which task in the company regularly takes longer than it should, and who is affected by it.
What you do not get
No recommendation for a tool before the process is clarified. No presentation about AI trends. And no assurance that a project pays off as long as the figures do not support it.
Or check for yourself first where you stand. The quick check on the overview page takes two minutes.