AI at your side. Never in your place.
We use AI to serve learning — never to stand in for it. We turn what is happening in the world into a reason to read, think and write in English — and we show families the evidence, not a score we invented.
What a lesson is
A short original story built from current science, maths and world events, six questions, one written explanation, and a companion who comes along. Lessons and questions stay in English in every interface language — that is the point.
What AI does here, and what it never does
The model builds the stories, drafts a note the teacher can use or ignore, and pre-scores the writing. It never scores a child’s ability on its own: a teacher confirms every score and signs every note before a family sees it.
How reviews work
A teacher reads the learner’s own words in a compact queue, adjusts the four language scores where the model got it wrong, and signs one honest note. The note reaches the learner’s Review page and the parent report the same day.
What we store about a child
A nickname, a year level, a consent date and a one-way PIN hash. No real name, birthday, school, photo, child email or phone. We never train a model on a child’s work or sell a family’s data.
Six companions
Anto, Candi, Mimi, Lumi, Nimo and Ria each host a kind of story and appear in feedback, badges and review reminders — so the interface reads emotionally, not only in words.
What we have not done yet
Speech assessment is not connected, and the library is young: new stories arrive every week, and a lesson stays marked “in review” until a teacher has accepted it. We would rather say less than overstate.