StudyRouteStudyRoute
How we use AI

How AI takes part in your decision — and where its role ends

AI helps us process university, visa and money data faster than a human could by hand. But studying abroad is a high-stakes decision — the cost of an error is a lost year, lost money or a visa refusal. So we don't hide AI behind a 'smart consultation'; we show exactly where it entered your decision and who answers for the outcome. Seven questions — seven of our answers.

  1. 1

    What did AI summarise?

    AI condenses large sources — university rules, visa requirements, costs — into a short answer for you. We always show which source a fact came from: link, year, publisher. A summary without a source never reaches the answer — that is the Fact tag in the provenance block.

  2. 2

    Did AI shape the shortlist?

    Yes — AI ranks universities against your profile. That is its job. But it is an estimate, not truth: it carries the AI estimate tag and is never presented as guaranteed admission. Ranking is a starting point for a human, not a verdict.

  3. 3

    Did AI form the first impression?

    It might — and that is exactly what matters to see. So AI estimates are visually separated from facts: you see at once where the verifiable ends and the model's judgement begins. A first impression you can't trace is dangerous; a traceable one is fine.

  4. 4

    Did anyone check the AI output?

    On a paid consultation — yes. The final route is approved by a live advisor; that is the Human decides tag. On the free quiz there is no such gate, and we say so honestly — there you get a bearing, not an approved decision.

  5. 5

    What evidence decided it?

    Only verified data for the five destinations we maintain by hand (Germany, the Netherlands, Estonia, Latvia, Lithuania). AI does not 'supplement from memory': if a fact isn't in our data, the answer carries the No data tag instead of an invented average.

  6. 6

    Where did human expert judgement enter?

    Where data isn't enough for a single answer: the priority between risk and speed, reading a borderline status, choosing a backup route. These are human decisions — and we tag them, rather than passing them off as the model's output.

  7. 7

    Who answers for the outcome?

    A human. AI lays out the options, an advisor checks them, the decision is yours — with our help. We don't hide behind 'the algorithm said so': every recommendation has an author.

AI lays out your options. A human checks them. The decision is yours.

And under every key recommendation — a provenance block: what is a fact, what is an AI estimate, where a human decides, what we don't know. Not because we're required to. Because what's at stake is your year.