How I assess business problems, distinguish useful architecture from unnecessary tooling, and why my positions come from building, not theory.
In most SMEs, quotes are built from email, PDF and tribal knowledge. What an end-to-end quote process looks like, what can be automated – and what deliberately stays with people.
Timesheets arrive as messy spreadsheets, the firm's software expects clean import files for BMD or RZL. Why the gap in between eats firm capacity – and how a checking overlay closes it without replacing the software.
Most mid-sized companies learn about market moves by accident – at the trade fair, from a customer, too late. How automated competitor monitoring works and why the briefing must be interpreted, not just collected.
A knowledge chat is useful only when the sources are right. Why the real work happens before the chat and how operating knowledge becomes a usable system instead of a bottleneck.
Automation pays off only in clear, recurring and stable workflows. When workflow automation helps, when it is overengineering and why the process must be right first.
A useful AI module does not replace the company's decision. It structures, prepares and leaves the human review exactly where responsibility is created.
The first AI case decides whether trust grows or the topic is burned. Five criteria show which start carries weight and why the most spectacular case is rarely the right beginning.
Individual prompts save time, but they do not yet build a system. The leverage appears when context, repeatable workflows and clear responsibility come together.
AI is not a yes-or-no question. What matters is which data is processed, where the system runs and which operating model fits the protection need.
Employees are already using AI. Management's task is not a ban, but a safe frame: approved tools, clear data rules and practical training.
Fragmented tool landscapes cost more than licenses. Why duplicate data entry, broken workflows and missing single sources of truth slow growth, and why integration is often better than the next tool purchase.
Not everything belongs in the cloud, but not everything belongs on the server in the building either. What matters is who can operate the system more reliably and who controls the data.
These positions show how I think. The projects show how I build. If you want to know whether that fits your situation, let's talk.
Marc Schraepler von Gerlach
Digital Architect for the growing Mittelstand. I decide what genuinely moves you forward and build it myself, from strategy to a running solution.
© 2026 Marc Schraepler von Gerlach
