EdTech Realist

Substack AI Policy AI Check Renata EdTech Contact

I. The problem, in one sentence.

Your school is
"working on AI."

I am a school owner who fell too deep into AI. I help heads of school cut through the noise and decide what actually belongs in the building.

Book a call Check if you need my help

Budapest.
est. 2026

scroll ↓

AI for teachers is a must.
AI for kids is the debate.

Start here.

Most schools come to me trying to write an AI policy.

Almost every draft I read is missing the same few paragraphs. Six questions, three minutes, and you will know which ones are missing from yours, plus what to write instead. No cost, no call required.

Take the AI Policy Starter

the policy is the easy half. the decisions underneath it are the check.

II. The offer.

The AI
Decision
Map.

fixed scope. fixed price. three weeks.

Someone got assigned. A tool got bought. A policy got written. And the real decisions are still sitting on your desk. Money. Parents. What happens in the classroom on Monday morning.

Some schools need a policy. Some need the whole map. Most start by finding out which. In three weeks every unresolved AI question your school is facing lands in one place, with one owner per decision and an order to take them in.

I.Week one

Evidence.

I talk to the people who actually know. You, the person you assigned to handle AI, a handful of teachers, IT. I audit what got bought, what gets used, what the policy says, and what students really do.

II.Week two

The map.

Everything lands on one page. Decisions that are yours alone. Decisions you can delegate. Safety and policy gaps nobody flagged. The two or three quick wins hiding in tools you already pay for.

III.Week three

The decision session.

I present the map live to you and your leadership. You leave knowing what to decide, in what sequence, and who owns what. No sixty-page report. One map, one sequence, one owner per decision.

Founding-client pricing — schools

The AI Decision Map

€6,500

fixed · three weeks

Every AI decision the school faces, who owns each, in what sequence, cost of postponing. Fixed scope, fixed price. Leadership-level diagnostic, costs meaningfully less than the year of implementation it opens.

The Build

From €25,000

ninety days · scoped by the map

Doing what the map found. Named deliverables, defined end, no open-ended retainer. Scope is set on the last page of your map, not in a sales call.

No Build starts without a decision map. If your school already has one that holds up under review, we say so and move straight to execution. If not, the Map comes first.

Half up front, half on delivery. The second half is due only once your decision room has done its job: every priority question decided, assigned, or in a bounded test. Closure is the deliverable, not a report.

Paid to check, not to cheer. The map sits on your side of the table, testing confident claims against evidence. Your team does not need to be convinced first. Only you do.

Bring your messiest AI question

Thirty minutes. If I am not the right person, I will tell you in the first ten.

III. The map in motion.

Every unresolved AI question leaves with a next move.

TEST One enters a bounded test.
WATCH One stays under watch.
STOP One expensive mistake stops.
SCALE One useful practice scales.

Question named→Evidence graded→Owner assigned→Decision made→Review date set

IV. What you actually get.

A decision folder built from your school. Six working records. Zero theatre.

This is the operating system left behind after three weeks. Select a folder to see the shape of the work. The evidence stays deliberately obscured, because the real version is built inside your organisation.

01 / 06 · WORKING ARTIFACT

Executive Decision Map

One page. Every priority decision, its state, owner, dependency and next review date.

CLIENT ONLY

WATCHTeacher-facing AI workflowOwner assigned

STOPStudent-facing use caseClosed

SCALEProven staff practiceReview set

V. The three weeks.

What changes in three weeks.

The map is the mechanism. This is what you are actually buying.

When we enter

✕AI tools scattered across departments
?Teachers using personal accounts
✕Renewals continuing by default
?Nobody certain who can approve what
✕The same questions reopened in every meeting

When the map leaves

✓Priority questions visible in one place
✓Every decision has one named owner
✓Tests have boundaries and expiry dates
✓Stop and scale decisions documented
✓The board sees one accountable view

Not sure where to start?

See where your school actually stands.

Twelve leadership questions, ten minutes, no technology knowledge needed. You get your two weakest areas and a thirty-day move for each. Free.

Check if you need The Map

VI. The founding ten.

Bring the AI question your team keeps reopening.

Thirty minutes. We will identify where it belongs: stop, watch, test, adopt, scale or retire. If I am not the right person, I will tell you in the first ten.

Prefer to watch first? Send me the first cohort's results.

VII. Field notes.

What the work actually looks like.

hover to peek. click to open the full frame.

VIII. Who is behind this.

Renata Kecskes.

I founded an accredited private high school in Budapest and brought AI into the curriculum while most schools were still debating phone bans. The school still runs without me in the building.

I have made payroll. I have handled the angry parent at 9pm. I have hired the wrong head of department and had to fix it.

I run AI training programs for teachers and corporate teams from Mexico to Dubai. I write EdTech Realist.

My starting point never changes. AI for teachers is a must. AI for kids is the debate.

SAS Budapest FutureTour LinkedIn

IX. The writing.