Services Case Studies About Insights Contact Let's talk
DE|EN
SERVICE · AI & AUTOMATION

AI & Automation — pragmatic, not buzzword.

AI deployed pragmatically — not as hype, but as a concrete tool for measurable efficiency gains in your processes and projects. Self-hosted where it makes sense, without unnecessary cloud dependency.

Discuss AI use case See scope of services
SCOPE OF SERVICES

Five areas where AI
makes a measurable difference.

No buzzword bingo — concrete tools, concrete use cases, always directly with me.

01
AI Strategy & Roadmap
Where does AI genuinely pay off in your processes, and where is it just hype? An honest prioritisation instead of blanket promises.
02
LLM Integration into Your Tools
Embedding language models where they save real effort — from document analysis to script generation.
03
Process Automation with AI
Automating recurring workflows, with human oversight exactly where it matters.
04
AI in Plant Engineering
P&ID and document analysis, AI-assisted evaluation of engineering documents and transfer into your COMOS structures.
05
Self-hosted AI Infrastructure
Where data protection or cost control demand it: AI infrastructure that stays in your hands, not a third-party cloud.
CONCRETE EXAMPLE

How a document
analysis project runs.

A typical process when P&ID drawings or specifications need to be evaluated with AI — from the first test run to production use.

01
Test run with real documents
We start with 5–10 typical documents from your own archive — not demo data. That immediately shows where the actual stumbling blocks are.
02
Define prompt & structure
Together we define which fields need to be extracted and in what format they're transferred into your systems (e.g. COMOS).
03
Validation by your team
The AI delivers a draft, your team spot-checks the first runs — trust comes from verification, not from promises.
04
Production use
Once validated, the evaluation runs in everyday operation — with clearly defined points where a human still reviews it.
HOW IT COMES TOGETHER

COMOS meets
artificial intelligence.

I develop script generation and code review for VBScript and .NET automations faster with AI support — less time spent, same quality. Engineering documents, specifications and P&IDs get evaluated with AI support and transferred into COMOS structures. And I also advise on how your plant engineering team can integrate AI sensibly — concrete tools, concrete use cases.

More on COMOS consulting
FREQUENTLY ASKED

Before you
reach out.

Q1
Does our data stay in-house?
Where it makes sense, I use self-hosted models (e.g. via Ollama). Where cloud models like Claude or GPT are needed, we discuss beforehand exactly what data leaves the building — and whether an on-premise alternative is enough.
Q2
Which models do you use?
Depending on the use case: Claude, GPT, or self-hosted models via Ollama — never a model out of principle, but whatever fits your requirements around cost, privacy and accuracy.
Q3
What does a first test run cost?
The free initial call clarifies scope and feasibility. A small test run with real documents is often possible within days, before deciding on a larger project.
Q4
What if the AI makes mistakes?
That's why a human stays in the loop at critical points — the AI delivers a draft, not an automatic final decision on critical data.
NO OBLIGATION

Want to check AI use
in your team?

Whether it's a concrete automation or a first AI strategy — tell me what's ahead, and I'll give you an honest answer on whether and how I can help.

Discuss project