Making EuraStudy
One Platform, Four National Exams
The Austrian Matura and the German Abitur are live; the French Baccalauréat and Spanish Selectividad are on the waitlist. The hard part was never the content — it was deciding what four exams could share without flattening any of them.
A study platform that promises to prepare a student for a national school-leaving exam is making a quiet, enormous claim. It is saying: we understand how your country decides whether you are ready. That understanding is not a feature you can copy from one border to the next. The Austrian Matura, the German Abitur, the French Baccalauréat and the Spanish Selectividad are not four translations of the same test. They are four philosophies about what a finished education looks like — and four very different ideas about how to grade it.
Today, two of those exams are live on EuraStudy: the Matura and the Abitur. The Baccalauréat and the Selectividad are on the waitlist, in active build behind the same machinery. This is the story of the decision that had to come first — before a single question was written: deciding what four exams could honestly share, and what each one had to be allowed to keep.
The temptation of the average
The cheapest way to serve four curricula is to serve none of them. You build a generic "high-school maths" course, dress it in four flags, and let the marketing imply more depth than the content holds. It is a tempting shortcut, and it is everywhere.
It fails the moment a real student opens it. An Austrian arrives expecting the Kompetenzmodell and the SRDP operator vocabulary that the Matura is actually graded against. A German expects the KMK operators, the eA/LK track distinction, and marking guidance that allocates points the way a real Erwartungshorizont does. A French student expects the barème, the correcteur's logic, the specific weight a dissertation carries. A Spanish student expects the structure of the EBAU, its optional blocks, its regional variation. None of them will accept an average. An average is a fifth exam that no country administers.
So the first principle was stated as a prohibition: never grade a student against a curriculum their country does not use. A question is not just a prompt and an answer. It is a prompt, an answer, and the grading culture that gives the answer its meaning. Strip that culture away and you have not simplified the product — you have falsified it.
A spine, not a blender
If averaging is out, the opposite temptation appears: build four separate products, four codebases, four teams, four slowly diverging definitions of "a topic." That path is honest and unaffordable. It also throws away the one true thing the four exams have in common — they are all attempts to certify a serious secondary education, and they overlap enormously in what they teach even as they diverge in how they assess it.
The resolution was to find the layer that is genuinely shared and make only that layer common. We call it the spine.
The spine is a single graph: subjects hold topics, topics hold questions and notes, and learning activity flows along the same edges no matter which exam you sit. A topic on derivatives is a topic on derivatives whether the student will be examined in Vienna or Munich; the underlying mathematics does not respect borders. What changes is everything above the spine — the operator vocabulary, the point allocations, the marking language, the very words on the chrome. Each curriculum lights the same constellation in its own colour.
Four exams could share a skeleton. What none of them could share was a soul. The craft was keeping the skeleton invisible and the soul intact.
This is why the platform can be one publication and four exams at once. The Matura and the Abitur run on the identical question-card primitive, the identical notes renderer, the identical figure engine — and yet an Austrian sees Matura operators and a German sees KMK operators, because the curriculum-specific layer is data riding on a neutral structure, never a fork of the structure itself. When the Baccalauréat and the Selectividad come off the waitlist, they do not arrive as new apps. They arrive as new lights on a constellation that is already standing.
Where honesty lives
A shared spine is only safe if it carries honesty across every border without leaking. The load-bearing rule of the whole platform is small enough to print on a card: a number on screen is either a real assessment or it does not exist. No fabricated scores, no invented confidence percentages, no placeholder mastery bars dressed up as data. Where something is unknown, the platform says so.
That rule is harder to keep across four grading cultures than within one, because each culture allocates points differently and each is an opportunity to quietly cheat. The defence is provenance. Every question in the verified bank — roughly three thousand of them today — carries a ledger: where it came from, who reviewed it, how its points are allocated, and against which curriculum's operators it is graded.
The ledger is what lets a question move along the spine without losing its truth. A Matura question's point allocation is the Matura's, not a translated approximation of it; an Abitur question's marking criteria sum to the points the Abitur actually awards, or the question is dropped. When AI assists — explaining a step, grading a photographed answer — it is labelled as assistance, never laundered into the verified bank. Honesty does not survive translation by accident. It survives because every claim drags its evidence behind it, in every language.
One publication, four readers
The last decision was the quietest, and the one a reader feels without being able to name. Four exams, served from one spine, with honest provenance throughout — and still, the thing has to feel like a single, calm publication rather than a switchboard.
That is the work of the design language. One serif voice for every headline. One monospaced voice for every caption and code. A strict palette of editorial inks, with a country's flag colour permitted only as a small accent and never on the working chrome — so the page reads as the same magazine in Vienna, Munich, Paris or Madrid, with the curriculum's identity carried by a typographic index code rather than a coat of paint. Diagrams are drawn by a single figure engine to one specification, never pasted in as decoration. Motion is considered and reduced-motion-safe; nothing animates to impress.
The result is a platform that asks the student to think about their exam and nothing else. The machinery underneath — the shared spine, the provenance ledger, the four grading cultures held apart so none of them blurs — is supposed to disappear. A French student arriving next year should never feel they joined a German product wearing a tricolore. They should feel they opened a publication that was, somehow, already fluent in the Baccalauréat.
Two exams are live and two are waiting, but the decision underneath all four is the same: honour each country exactly, and let nothing in the structure betray that you are doing it four times at once. The calm on the surface is the proof the structure held.
Start preparing with EuraStudy
Join the waitlist to be first in when your curriculum opens.
Adaptive Practice and Its Limits
Adaptivity is the most over-promised word in educational technology. Two effects in the learning-science record are real and worth building on; almost everything sold above them is decoration.