How it works

One lesson, four moves.

Every lesson follows the same loop. We tell you up-front so you and your child know exactly what to expect — and so you can spot when the AI is doing its job.

1

The lesson opens with a question, not a video.

The AI introduces the concept by asking what your child already knows. "If you have 12 cookies and want to share with 4 friends, how many does each friend get?" The first move is always the student's, not the lesson's.

Why: passive video creates the illusion of learning. Active recall creates actual learning.

2

The AI asks, nudges, and explains by analogy — never gives the answer.

When your child gets stuck, the AI doesn't drop the answer in their lap. It asks a smaller question. It rephrases. It connects the concept to something they already understand. This is how a good private tutor works. It's slow on purpose.

Why: knowing how to find the answer is more valuable than knowing the answer.

3

The mastery gate decides whether they advance.

At the end of every lesson there's a checkpoint the AI evaluates. Sometimes it's a written response, sometimes a verbal explanation, sometimes a scenario problem. The AI grades it the way a good teacher would: looking for genuine understanding, not memorized phrases.

Why: completing a lesson and understanding it are different things. Most platforms treat them as the same.

4

If they're not ready, they hear "not yet" — never "fail."

When a student doesn't pass the mastery gate, the AI loops them back through the parts they got wrong, with a different angle. There's no grade, no shame, no "F" on a transcript. Just another loop until the work clicks. Some students breeze through. Some take three passes. Both outcomes are fine.

Why: shame breaks the willingness to keep trying. We refuse to use it as a teaching tool.

What a mastery gate looks like

Three formats, one purpose.

The AI picks the right format for the lesson. The student sees one of these:

Written

Explain it back.

"In your own words, why does multiplying two negative numbers give a positive?" The AI evaluates whether the explanation shows the actual mechanism, not just the rule.

Scenario

Apply it.

"Your friend is short USD 12 to make rent. You have USD 40 saved. They'll pay you back in 2 weeks with USD 5 interest. Walk through whether you should lend it." Real-world reasoning, not just arithmetic.

Oral

Talk it out.

The student records a 90-second voice answer to a prompt. The AI transcribes + evaluates. We use this for English, history, and life-skills lessons where comfort speaking is part of the skill.

A real moment from a real lesson

How the AI handles a stuck student.

Tutor
If a triangle has a right angle and one leg is 3 and the other is 4, what's the hypotenuse?
Student
Uh, 7?
Tutor
Close in addition — but the hypotenuse rule is different. Do you remember what we squared?
Student
3 squared is 9 and 4 squared is 16…
Tutor
Right. So now what?
Student
Add them. 25.
Tutor
And then?
Student
Square root. So 5.
Tutor
✓ 5 it is. You used the rule cleanly. Want one more like this, or skip ahead?

Notice what the AI didn'tdo: it didn't say "the answer is 5." It walked the student to 5 themselves. That's the move that makes mastery stick.

See it for yourself.

14-day free trial. No card required. Add students after sign-in.