Agent-Centric Work | Hybrid Cognition

The Challenge

Your team already works with AI. Maybe even intensively. And yet the productivity gain falls short of what you expected.

Individuals achieve impressive results, but the knowledge stays with them
Experiments sprawl in every direction, but no shared practice emerges
Training sessions create excitement that fades within two weeks
Concerns about data protection and IT security block the genuinely interesting use cases

The Solution

More training and more individual experiments will not close the gap. What's needed is an energising, shared experience that gives your team a common language and turns productive work with AI agents into a shared goal.

My method has three steps:

1

Discovery

2–3 weeks | With management

Before your team spends a day with AI agents, we set the foundations:

What is already working, where are the biggest levers?
Which decisions about tools, data protection and security need to be firmed up?
What will be the focus of the workshop challenge?

2

Workshop

1 day | On site | 8–12 participants

Paintball for the brain.

Two teams. One AI agent per person. A real challenge from your business. Under time pressure, the teams use their agents to build something ambitious. Then they switch sides and pull the other team's work apart. The solutions that looked airtight collapse in minutes. Everyone learns why.

Build. Break. Reflect.

The workshop creates what no training session can: a shared language and a shared judgment about where AI agents deliver and where they fail. Up close, under pressure, with energy and ambition.

3

Embedding

4–6 weeks | Weekly cohort

The energy from the workshop becomes a shared way of working through daily practice.

In weekly sessions we talk through successes, failures and open questions. Participants learn from each other. They become AI champions who carry the knowledge back into the organisation.

The goal is independence: the point at which your organisation teaches itself and no longer needs me.

The Change

A shared vision and language

Your team understands where the journey with AI is heading and speaks the same language. No more isolated experiments, but shared practice.

Judgment instead of opinions

Your people have experienced first-hand where AI agents deliver and where they fail. Future decisions about AI use rest on experience, not on hype or hope.

An organisation that teaches itself

Participants who use AI agents productively every day and actively pass on what they learn. Structures in which experiments, successes and insights flow back into the team. The organisation teaches itself.

Testimonials

Martin gave our team outstanding support in introducing AI agents. His approach convinced us from the start: instead of creating dependencies, his goal was always to enable us to deliver our projects ourselves. His support landed exactly where it was needed — and that is precisely why we work independently and confidently today with the solutions we built together.

The workshop was inspiring and highly practical — we didn’t just talk about AI, we actually built things. My biggest aha-moment was seeing how confidently my AI agent executes its job once you give it a detailed and well-crafted skill. I implemented two concrete use cases: an agent that automatically handles incoming partner requests on revenue analyses, visualizes the data and sends it back to the partner via email — while simultaneously notifying me via Telegram. And a second agent that proactively contacts partners every week with current revenue developments and concrete improvement suggestions. Exactly what I was looking for.

In one morning with Martin we built an agent that compiles the weekly competitor analysis from web sources and press releases, publishes it as an HTML page on our web server, and sends me the summary via Slack and email. After 27 years at bergfex I have watched many waves of automation come and go — what convinces me about Martin’s approach: he starts with the question of which use case actually makes sense, before anyone touches the build.

The workshop with Martin was simply brilliant. In just three hours we developed a bot that connects my email, Slack, Gmail and Todoist. Since then it automatically organises all my to-dos from various channels every day – an enormous relief in my daily work.

Martin gave me the foundation to understand and work effectively with OpenClaw. Beyond technical guidance, he encouraged me to explore its full potential and apply it creatively to my business needs. Thanks to this, I built daily and weekly reporting tools that provide all my key KPIs across devices I already use.

Martin is a real learning accelerator. After our one-on-one coaching I use Claude in ways that were closed to me before – not as a gimmick, but as a genuine tool in my daily work. I have gained a lot of time. What sets Martin apart: he really listens. He starts from my existing working methods and tries to keep as much as possible the way it was – and to evolve from there. Not a fresh start on a greenfield, but an evolution. My questions about security and data protection were also taken seriously. Anyone who really wants to use Claude – not just on the surface, but the way advanced users do – is in the right place with Martin.

Martin Schröcker

I run my company with three AI agents as colleagues. They research, analyse, code, and challenge my thinking. The adrenaline-charged sprint with agents has become my daily work.

After fifteen years as an engineer, strategy consultant and CTO, AI gripped me three years ago. Today I am convinced: in the AI era, the people and organisations that win are the ones who have learned to put AI at the centre of their work without giving up control. Guiding this shift is what drives me every day.

LinkedIn →

Frequently asked questions

«Does this fit our situation?»

This format is built for organisations that already work with AI, perhaps intensively, but sense that the productivity gain falls short of expectations. Knowledge stays with individuals, experiments don't add up to shared practice. The workshop creates exactly that shared foundation.

Even if your team is still at the start: if you are ready to take AI seriously and your team needs an intense experience to get started together, the format works just as well. Get in touch, and in the first conversation we'll clarify whether it fits.

«Does this fit our industry?»

Every workshop is adapted to your context. Your industry, your workflows, your strategic pressure points. Through challenges drawn from their own work environment, participants learn to judge when they can trust AI agents, when they need to verify, and when they should override.

The capacity to know when to trust AI, when to verify, and when to override is universal. The application is specific to you.

«Our team will be sceptical.»

Sceptical teams are the best participants. The format is competitive, not a training session. No one has to believe anything; the point is to win. The sceptics turn into the most engaged participants.

«What about data protection and IT security?»

In the Discovery phase we discuss the state of these topics and, where needed, firm up your decisions. Which data may be processed with which tools, which approvals are needed, how do you protect yourself?

My experience: data protection risks are often overestimated, while the operational risks of the tools are underestimated. Both can be addressed well with the right decisions.

«Which tools do you use?»

In my own work I mainly use Claude Code, Claude Cowork and OpenClaw. The methods behind them work with any AI agent. We adapt the workshop to the tool stack your team already uses or wants to use.

«What does it cost?»

Every engagement is tailored to your situation. In the first conversation we'll clarify together which format fits and discuss the investment.

The next leap in productivity won't come from a better tool. It will come from a new way of working.