Skip to content

The framework
before the decision.

Topics that give orientation before it comes to concrete steps. From our workshops for leadership teams, presented neutrally and without claiming to be complete.

What do we go by?

Strategy triangle

Responsibility, independence and pace in balance.

Three demands that limit each other. Those who only bet on pace lose control. Those who only play it safe make no progress.

Every company finds its own balance. The triangle helps to choose it deliberately.

“All three count. The weighting is up to you.”

EthicsPeople at the centre
SovereigntyData and processes in your own hands
AgilityLearn fast, scale deliberately

In what order?

Levels

No third floor without the second.

AI builds on what lies beneath it. Where data is not captured and processes are not mapped, AI lacks the foundation it needs to work with.

The levels say nothing about ambition. They say in which order effort pays off.

“The first level is rarely finished. That is exactly where it pays off first.”

What has to be in place first?

Three goals

What comes before any technology.

Three goals that sit on the lower levels. It starts with simplifying. The examples show what everyday work looks like when one of them is missing.

“None of these three problems is an AI problem. All three prevent AI.”

Goal 01

Simplify

Leave out what nobody needs. Fewer exceptions, fewer special cases.

The same task, many templates, depending on the person.

Goal 02

Structure

Data in a clean, machine-readable form. Only what is structured can be evaluated.

Customer data in several systems, none in sync.

Goal 03

Standardise

Processes that are repeatable. What everyone does differently cannot be automated.

Every quote is created a different way.

Where do we start?

Fields of action

Where AI can become effective in a company.

Four fields as a frame of reference. The fourth is the foundation for the other three.

Applications

Extend existing applications with AI functions or develop new ones.

  • Dashboards
  • Forms
  • Chatbots
  • Assistants
  • Search in your own knowledge

Automation

Automate recurring processes fully or partly.

  • Receipts
  • Approvals
  • Media breaks
  • Interfaces
  • Notifications

Agents

Systems that take care of several steps on their own. Based on your company data.

  • Research
  • Knowledge questions
  • Evaluations
  • Decision papers

Infrastructure

Where models run and where data lives. The foundation for the other three fields.

  • Model operation
  • Rights and roles
  • Data storage
  • Location

“Every proposal has a field. That makes it comparable.”

Where do we stand?

Maturity

With every level the topic gets broader and touches more of the company.

Observer

following, not in use

  • Ethics: No framework
  • Sovereignty: Data situation unclear
  • Agility: Waiting

Beginner

first tools, trying out

  • Ethics: First guidelines
  • Sovereignty: Risks known
  • Agility: Individual experiments

User

established in processes

  • Ethics: Guideline in place
  • Sovereignty: Jurisdiction partly settled
  • Agility: Targeted projects

Shaper

a natural part of how work runs

  • Ethics: Responsibility lived
  • Sovereignty: Full data control
  • Agility: Continuous scaling

“A position, not a grade.”

Four levels for the state of AI use, in your own area or across the whole company.

What do we measure every decision against?

“We act responsibly, stay independent and still move forward fast.”
  • Ethics
  • Sovereignty
  • Agility

And in your own company?

In the Discovery, this becomes your order.

Every starting point in the blueprint is assigned to a level and a field of action, with its prerequisites and with a view to the three corners. The framework stays the same; the answers come from your company.

How a Discovery works
Marcel Bauer
The decision stays with you. I make sure everyone is talking about the same thing.
Marcel BauerConcept and workshops · 25 years in conversation with executives

Request a workshop

These questions, discussed
with your leadership team.

A workshop that classifies instead of judging. Your team gains a shared vocabulary, sees examples from practice and takes away questions only they can answer.

  • Workshop
  • in person at your site