We didn't build LAURA to make grades faster. We built it to give every child a fair chance.
LAURA is the marking assistant for teachers, and that is precisely why it is a social project. Because every hour a teacher does not spend on a pile of exams is an hour for the children who need it most. And every child who first understands the material in their own language no longer simply slips through the cracks.
We believe that technology must serve people, not the other way around. That is why one rule applies with us, a rule we never soften: LAURA proposes, the teacher decides.
Where you come from must not decide where you can go.
In hardly any country does school success depend so strongly on the family home as it does here. Whoever gets help with homework at home, whose parents know the technical terms and can pay for a private tutor, has the advantage. Whoever is new to the country and is still learning the language of instruction loses touch, even though the child is often just as clever and hard-working as the others.
That is why, with every teacher subscription, we develop a learning app for pupils. It prepares them for tests, class assignments and exams, and translates the material into their heritage language. A child first understands the content in the language they think in, and then becomes confident in the language of the country. The app does not grade, it explains and practises along the same curriculum model that LAURA's marking uses.
Preparation, no pressure
The app practises and explains, it issues no grade. So a child can practise without fear, as often as they need to.
Learning in two languages
Task and explanation in the heritage language, answer and technical terms in the language of the country. A bridge, not a replacement.
Time for those who struggle most
Because LAURA takes the marking pile, the teacher keeps time for the children who need support, and for special talents.
Languages from crisis and heritage regions
We start with the languages most urgently needed in everyday German, French or British schools. The order is a proposal and will be prioritised with pilot schools according to real need. Does your class have other needs? Tell us, and those languages move to the front.
Arabic
Syria, Iraq, MENA region
Turkish
large community, a bridge
Russian
post-Soviet space
Ukrainian
refugee movement since 2022
Farsi / Dari
Iran, Afghanistan
Pashto
Afghanistan
Kurdish
Syria, Iraq, Turkey, Iran
Tigrinya
Eritrea, Ethiopia
Somali
Horn of Africa
Urdu
Pakistan, Afghan region
We add further languages according to real need.
Access staggered by social need
It is precisely the children with the greatest need who should be able to learn for free. Education must not fail because of the wallet, which is why access is deliberately open at the bottom.
5 voucher codes per subscription
With your subscription you receive five voucher codes for the children in your class who have it hardest. These pupils get free access to the learning app, without anyone at home having to ask whether the money is enough.
Free access for socially disadvantaged regions
Schools in socially disadvantaged regions and the teachers who work there receive unlimited access. This way integration in Germany, France or the United Kingdom can succeed, and not fail because of language or a rigid system.
Strict data protection: the app is data-thrifty, requires no real-name obligation for pupil use and passes nothing on to third parties. The heritage language is a voluntary detail.
We give teachers their time back, because otherwise the profession loses them.
You know the pile: thirty exams on a Friday evening, the weekend already half planned away, and on Monday the next one is waiting. Marking is the work no one sees, but the work that costs the most energy. It is precisely this invisible burden that drives many out of the profession, while teachers are missing everywhere. A vicious circle: where people are missing, classes grow larger, and the pile grows with them.
LAURA takes the dull routine off your hands. It reads the handwriting, matches it against the rubric and makes an assessment proposal with a rationale. You check, you correct where you disagree, and you release. Hours of working through become a focused review. The pedagogical decision stays a hundred percent yours, only the drudgery becomes lighter.
And the time regained works hardest where it is most lacking: at schools in tough areas, in large classes, with the children who need individual attention. This way the profession becomes attractive again, and the quality of education depends less on which neighbourhood a school stands in.
- Fewer evenings and weekends over piles of exams, more time for teaching and support.
- Fairer, more consistent marking, because tired eyes at midnight correct differently than in the morning.
- A profession that is again what it was chosen for: working with children, not with mountains of paper.
Children's data does not belong in a foreign cloud.
An exam is not an anonymous data set. It carries a child's name, their handwriting, their thoughts, sometimes also their weaknesses. Such data is among the most sensitive there is, and it must not become raw material for corporations on another continent. We treat it the way we would want it treated for our own children.
That is why LAURA is built European and data-thrifty from the ground up. A large part of the processing runs locally, directly with you or on a European server under our control. The cloud never sees real names, only pseudonym IDs; the real-name mapping stays local and encrypted. Cloud AI is used only deliberately and only for contested passages, not as the default.
It is also a question of sovereignty. European schools should not depend on what a provider overseas intends to do with their data, or on what its pricing model will cost tomorrow. AI sovereignty means: control stays here, and the decision is made by a human, never by the AI alone.
Local first
The bulk runs on your hardware or an EU server, not in a foreign cloud.
Pseudonyms only
The cloud sees no real names. The mapping stays local and encrypted.
Built to the GDPR
Data-thrifty, with consent, with data processing agreements. Nothing goes secretly to third parties.
A human decides
LAURA proposes, the teacher decides. No automatic grade output.
A grade must be explainable. Always.
Many AI systems are a black box: the exam goes in the front, a number comes out the back, and no one can say why. In school that is unacceptable. Parents have a right to understand how their child's grade came about, and you have to be able to defend it in front of the class and in parent meetings.
That is why LAURA works in the open. Every proposal is tied to the rubric: you see criterion by criterion how many points were awarded and why. The grading key is disclosed, not hidden. And when LAURA is unsure, it says so: yellow and red mark the passages that need a second opinion. Uncertainty is never hidden, but made visible, so that you can take a look.
Every decision is logged. A tamper-proof audit trail records what LAURA proposed and what you decided. So in the end every grade stays traceable, verifiable and yours.
Assessment along the rubric
Visible point by point: which criterion was met and how. No number out of nowhere.
Uncertainty is shown
Green, yellow, red. Contested passages land with you for review, instead of quietly slipping through.
A gap-free audit trail
Proposal and decision are logged tamper-proof. Every grade stays evidenced.
Education is too important to leave to chance.
Whether a child keeps up should not depend on which language is spoken at home, how full the class is or how tired the teacher is on a Sunday evening. We are building LAURA for a school where no one simply slips through the cracks. If you share that stance, let us start together.
