/ ADVISORY
You are not short of AI initiatives, opinions or vendors. You are short of verdicts.
Three ways to work with me. One settles a single decision. One decides what in your portfolio deserves attention. One turns your direction into a strategy your organisation can actually run. Pick by the decision in front of you, not by size. None of them requires the others.
Your portfolio keeps growing because nobody wants to kill anything. That is not a capability problem and no framework fixes it. It ends when a named person makes a call and writes down what would reverse it.
/ EACH STANDS ALONE
One-to-one consultation
Portfolio verdict
AI Strategy engagement
/ CONSULTATION
One decision, one session, and you keep the decision.
A paid engagement in its own right. It is not an intake step, it is not credited against anything else, and it is not the free fit call.
/ PORTFOLIO VERDICT
Five days to decide what deserves attention. One verdict against every line on the list.
This is where the four verbs belong. They are the instrument used to judge the portfolio, not the name of the work.
By Friday every initiative sits under one of the four, with a named owner, a measure and the written condition that would reverse it. What deserves attention now, what waits, what needs evidence first, and what should end.
Two sessions, roughly half a day each, with the executive who owns the decision present. Sponsors join the second.
A stop decision made by an advisor has no authority and no survival rate. It comes back as a renamed project two quarters later.
/ AI STRATEGY
/ FLAGSHIP
Six weeks to create an AI strategy short enough to read, specific enough to own.
The strategy has no authority when the real owner leaves the room. So six weeks assigns it: priorities, ownership, measures, sequence and governance, each with a name against it rather than a function.
/ AFTERWARDS
/ COMPARE
Read the table and choose. Everything you need to pick between them is here.
The exclusions row is the important one. It is what each engagement refuses to do, so nobody buys the wrong thing and calls it a misunderstanding.
Consultation
- BEST USED WHEN
- One decision is blocking you and you want it settled properly this week.
- YOUR TIME
- One hour, plus a paragraph written beforehand.
- SCOPE
- One question. Nothing else is examined.
- YOU PROVIDE
- The question and the two options you are between.
- PIOTR DELIVERS
- A position, the reasoning, and the reversal condition, in writing.
- EXPLICITLY EXCLUDED
- Portfolio review, artefacts, follow-on work of any kind.
- COMMERCIAL
- €500
Portfolio verdict
- BEST USED WHEN
- The list is long, spend is real, and nothing has been stopped or scaled in a year.
- YOUR TIME
- Two half-days, decision-holder present. Sponsors join the second.
- SCOPE
- The whole portfolio, judged. No strategy written.
- YOU PROVIDE
- The real list, spend per line, sponsor names.
- PIOTR DELIVERS
- Verdict map, decision record, owners, stop conditions.
- EXPLICITLY EXCLUDED
- Writing the strategy, benefit cases, vendor selection, governance design.
- COMMERCIAL
- €8,500
AI Strategy
- BEST USED WHEN
- Direction exists but the organisation cannot act on it without you present.
- YOUR TIME
- A recurring session each week, plus owners in weeks 04 and 05.
- SCOPE
- Strategy, priorities, ownership, measures, sequence and governance.
- YOU PROVIDE
- Access to the exec team, the portfolio, and whatever passes for the current plan.
- PIOTR DELIVERS
- Six artefacts: strategy, sequence, ownership map, measures, cadence, decision log.
- EXPLICITLY EXCLUDED
- Implementation, build teams, model work, platform engineering, budget approval.
- COMMERCIAL
- From €50,000
TO ALL THREE IF
It is not a step in any of the engagements and it is not a sales call.
/ PROOF
This is where the judgment comes from. The organisations it was built inside.
No client quotes and no case studies: the engagements are private. What is checkable is where the work was done, and what gets published under my name every issue.
/ THE WORK WAS DONE






Read four issues before booking anything. If the reasoning does not hold up for you there, it will not hold up in a session, and you will have saved yourself the fee.
READ THE WRITING/ QUESTIONS
Four questions people ask before they commit money to this.
Anything else, write to me directly.
Do I have to do them in order?
No. They are three routes, not three stages. Most people who take the six weeks have never had a consultation, and plenty of consultations are the only thing that ever happens. Start where your decision is.
How much of my own time does this take?
More than you would like, and it is not delegable. One session for the consultation. Two half-days for the week. A weekly decision session for six weeks. If your calendar cannot hold that, the work will not survive contact with your organisation.
Does my data need to be in good shape first?
No. Every one of these engagements is about judgment, ownership and sequence, and all three can be decided on incomplete data. If the honest answer to a question is that nobody knows, that is a finding, and usually the most useful one.
Where does your work stop?
At the decision and the structure that carries it. I do not build, staff or run implementation, and I take no vendor commission. Everything discussed is confidential and no client is named without written permission. Where specialist help is needed I will say so rather than extend my own scope.
/ NEXT STEP