00
About00 / About
I have spent twenty years on the side of the table where the decision actually gets made. That is the only qualification that matters here.
Currently
Executive AI strategist and advisor to CEOs. One week to decide what deserves attention. Six weeks to put the AI strategy into operation. Based in Warsaw, working across Europe.
01
Where I stand now01 / Where I stand now
I advise CEOs. I also own data and AI decisions inside a company.
Current delivery keeps the advice useful. I stay close to budgets and the point where plans break.
Advisory
Founder advisory to CEOs and owners
Board-level work on the AI portfolio: what to fund, test, scale and stop, with owners, measures and stop conditions attached. This is the practice this site is for.
Operating
Chief Data Officer, software company
I own the data and AI agenda inside the business, where every priority competes for money and executive attention.
Writing
Natural + Artificial Intelligence
One argument every issue, written for the people who own the portfolio and have to defend it. The thinking is public before it is billable.
02
Operating contexts02 / Operating contexts
Selected roles, and what each one taught me.
Not a chronology. Five contexts that changed how I judge an initiative, each with what it actually taught.
Data and AI priorities compete for money and executive attention.
Governance starts with a named owner and a written stop condition.
GovernanceExecutive leadership and public representation of the strategy.
A strategy you cannot defend on a stage, in plain language, to people who did not write it, is not finished.
SubtractionBuilt, launched, and learned the price of timing.
Being right about the product and wrong about the moment costs the same as being wrong about both.
TimingBanking and energy. Regulated, large-scale, multi-year.
Scale punishes vagueness. An initiative with no named owner drifts quietly for four quarters.
ScalingOne-to-one work with senior leaders. Low frequency, by design.
Most portfolio problems are not analytical. Someone is avoiding a conversation.
Leadership03
Two stories03 / Two stories
Two experiences explain how I judge. One cost me money. One taught me to slow down.
I built a product people wanted, several years before they were able to use it.
The detail was right. The experience was right. Product-market fit was real, and it arrived on a market that had not yet built the rails underneath it. There was no measure in place that would have told us to stop, so we did not stop when we should have.
That is why STOP. is one of the four verdicts and not a footnote. I have paid for the version of this work where nobody owns the ending.
Read: I Burned €50 Million Building Apple Pay Before Apple →
Underwater, pushing harder burns the oxygen you need.
Panic tells you to push. The correct response is to relax, slow the heart rate and accept the discomfort for longer than feels reasonable. The people who do this well are not the strongest ones in the water.
Boardrooms under AI pressure behave the same way. The instinct is to add initiatives, announce a programme and show activity. The counterintuitive move — fewer things, named owners, a written condition for stopping — is the one that works.
Calm is not a temperament I market. It is a working method, and it is trainable.
Read: Panic Looks Like Productivity →04
How I work04 / How I work
Four commitments. You can hold me to each of them.
Values are only useful if they are falsifiable. Each one below states what it rules out.
Essential focus
Choose the few initiatives that deserve attention and stop the rest.
Responsible decisiveness
Make consequential choices through evidence, explicit ownership and stop conditions.
Practical usefulness
Produce things leaders can use on Monday: roadmaps, scorecards, decisions and cadences.
Human agency
Keep humans responsible for direction, judgment, taste and moral choices.
CALM + COURAGE = CLARITY