Skip to content

AI that works in your company’s daily routine

I help owners and managing directors of small and mid-sized companies use AI where it saves real time. Easy to understand, safe for your data and without a large-scale project.

Landscape of white cubes with a yellow line running through it as a clear path
Columns of cubes rising step by step, topped with yellow

What AI actually does for your business

Quotes in a fraction of the time

A request arrives by email with a drawing attached. AI extracts the data and drafts a quote using your prices and standard wording. Your sales team checks it and sends it out.

Company knowledge everyone can find

Manuals, service reports and contracts become searchable with a simple question. New employees find answers on their own instead of interrupting experienced colleagues.

Less manual data entry

Delivery notes and incoming invoices are read automatically and passed to your ERP system. Your team only handles the cases that really stand out.

Faster replies to customers

Incoming requests are sorted and replies are drafted, in other languages too if needed. The final decision stays with your people.

Get more out of the AI you already use

Many teams have tried ChatGPT or Copilot and been disappointed. Usually the way of working is the problem, and that can be changed quickly.

Context instead of a single sentence

Before: “Write a reply to the customer.”
Better: Include the complaint, an excerpt from the contract, the tone you want and how far you are willing to accommodate. The result needs hardly any rework.

Team assistants instead of individual chats

Before: Everyone writes their own prompts from scratch.
Better: A shared quoting assistant knows your products, templates and wording. Quality no longer depends on one person.

Connect instead of copy

Before: Text is copied and pasted back and forth between programs.
Better: AI accesses your inbox, file storage or CRM directly, with defined permissions.

Review with a system

Before: Results are either taken over unchecked or not used at all out of distrust.
Better: AI cites its sources, and a person always approves before anything is sent.

Scattered cubes that turn into a precise grid from left to right
A tall, messy stack of paper next to a flat, neat stack with a yellow cover sheet

Before and after: three typical cases

This is how work changes in companies of your size.

Note: These are example scenarios, not customer references. All times are assumed values for illustration.

Mechanical engineering, 40 employees

Before: Preparing a quote from a customer request takes about 45 minutes. The information is scattered across emails, drawings and old quotes.

After: AI delivers a draft within minutes. Sales reviews it in about 10 minutes and can answer more requests on the same day.

Skilled trades, 15 employees

Before: In the evening, the site manager spends about an hour typing up reports and measurements from handwritten notes.

After: He dictates into his phone on site. AI turns this into a structured report that reaches the office after a quick check.

Wholesale, 25 employees

Before: Two people type orders from PDF files into the inventory management system every day. This leads to errors and follow-up questions.

After: Orders are recognized and entered automatically. The team only handles unclear items.

AI then and AI now

If you checked a few years ago whether AI pays off, it is worth running the numbers again. The barrier to entry has dropped sharply.

Then

  • Custom models had to be trained on large amounts of data.
  • Projects took many months and required data science specialists.
  • Each system could do exactly one task.
  • Operation only through specialized software.

Now

  • Ready-made language models understand text, images, spreadsheets and speech.
  • A first pilot is often up and running within a few weeks.
  • AI agents carry out work steps across several programs, and people approve them.
  • You use them in plain language, without programming skills.
A single sphere on a pedestal next to a network of connected spheres
Several rings of cube walls surrounding a yellow sphere in the center

Your data stays under your control

Sensitive content and AI are not mutually exclusive if the rules are set in advance. This is my approach:

Classify first, then use

We define which data is public, internal or confidential, and which tool is allowed for each level.

The right tool for each level of protection

For many tasks, business plans with a data processing agreement, data storage in the EU and no use of your data for training are sufficient. We check these points for each provider. Particularly sensitive content can stay on a local model on your own premises.

Existing access rights still apply

Document search only shows each person what they are already allowed to open today. Payroll lists stay with HR.

People decide, training provides safeguards

Important steps are approved and logged in a traceable way. Since February 2025, the EU AI Act has also required sufficient AI literacy from everyone who uses AI in a company. That is why the right training is part of the package.

I clarify legal questions together with your data protection officer. My consulting does not replace legal advice.

Questions decision-makers often ask me

Large question mark made of many small spheres with a yellow sphere as the dot

Which task takes up most of your team’s time?

Tell me about it. In the initial consultation, we check whether AI really helps there and what the most sensible first step is.

Privacy Overview

This website uses cookies so that we can provide you with the best user experience possible. Cookie information is stored in your browser and performs functions such as recognising you when you return to our website and helping our team to understand which sections of the website you find most interesting and useful.