Schools move on before mastery.
Curricula march forward whether your child is ready. A shaky fractions unit in Grade 3 turns into a rocky ratios unit in Grade 5 — and the gap compounds. Nobody pauses to close it.
You pick what your child is learning. Algo School turns it into 10–20 minutes of daily practice that adapts to them — with a tutor that never gives the answer away, and proof you can see.
15 days free · cancel in one click · kids use aliases, never real names
Pico cheers the effort — the answer stays the kid's to find.
Algo School is school practice your child actually completes — run by you. Public school teaches the concepts; mastering them takes practice at home, and that's the part parents are left to figure out alone. Algo School gives you the plan, subject by subject: math is live today (aligned to Common Core, Grade 1 through Algebra 2), ELA is next, and more subjects follow. Pick the concepts your child is learning and it generates short daily practice sessions from a library of 99,079 verified questions. Every answer is checked, every session updates your child's mastery profile, and the next session adapts — harder where they're strong, focused where they struggle.
When your child gets stuck, a built-in AI tutor teaches the way a good teacher does: with questions, hints, and diagrams — never by giving the answer away (it's engineered so it can't, even when kids try). After every tutoring session, you get a short email: what they worked on, what clicked, what to ask at dinner. You stay the coach. Algo School is the assistant.
Homeschooling? This is the model the whole product is shaped around — see how Algo School fits a homeschool math program.
Four problems compound between kindergarten and middle school. None of them are your kid's fault. All four share the same root cause — practice that doesn't build the right thing, in a routine that screens are happy to fill instead.
Curricula march forward whether your child is ready. A shaky fractions unit in Grade 3 turns into a rocky ratios unit in Grade 5 — and the gap compounds. Nobody pauses to close it.
Most kids practice the way they brush their teeth: quickly, randomly, when reminded. No deliberate sequence, no idea which gaps matter most. A year of "doing math" later, no one can say what actually got built.
On most platforms you have no idea which videos your child watched, which questions they attempted, what they got wrong, or which concepts they're hiding from. Time spent looks like progress in the app. It usually isn't.
The iPad doesn't care. Without a 15-minute daily anchor, time leaks into games, YouTube, and reels. Kids fall behind not because they're lazy — but because their day has no learning-shaped hour in it.
Sign in with Google, set a parent PIN, add your child (an alias and a grade — never a real name), pick subjects, choose a buddy and a strategy (year plan, mastery, or topics) — about five minutes, once. After that the week runs itself: teach → practice → master, with each round informed by the last, and you always know what happened.
Choose what your kid is learning this week. You're the coach — Algo School is the assistant.
10–20 minutes. Questions adapt — harder where they're strong, focused where they struggle.
What clicked, what didn't, and what to ask at dinner. Not a "time spent" number.
11 math courses, every concept backed by verified practice questions before we let you select it. We list exactly what's ready — and we don't claim what we haven't shipped.
"Check your answer" does nothing for a child who is confidently applying the wrong rule. So every concept in Algo School carries the specific ways children go wrong on it: the mistaken rule, why it feels right to a child, the sentence you are likely to actually hear — and what to say back.
We publish the diagnosis half openly, before you sign up — read the write-up for a concept your child is stuck on and see whether it describes them. There are 50 of these up right now, and we're adding more.
Real math is not multiple choice. When the answer is a fraction, a power, a root or an inequality, your child types the real thing — and can work it out by hand first, on the screen, the way they would on paper.
Bloom's Taxonomy is the standard educators use to classify cognitive demand. We tag every question we generate by its level — from Remember (recall) to Create (novel problem-solving). Practice spans the whole pyramid as mastery builds, so kids aren't stuck at recall forever.
A normal session blends three difficulty tiers: 30% Introductory (warm-up), 50% Reinforcement (the meat of the session), 20% Challenge (push). When a concept's mastery dips, the mix shifts toward fundamentals (50/40/10); spaced-review concepts flip the other way (10/30/60, mostly challenge). And mastery per concept decides which concept comes next — so practice targets what your kid actually needs this week, not just what's next in a book.
We track mastery with a weighted recent-average score per concept, so one lucky run doesn't fake it and one bad day doesn't tank it. Concepts move through four states: Not Started → In Progress → Proficient → Mastered. Regress on later questions and the state drops back to Needs Review so we know to revisit it.
AI helps us build our question bank — it never gets to improvise in front of your child. Math questions come from 177 deterministic engines we wrote (a fraction engine knows how to make a "compare ¾ and ⅔" question correctly, every time), and each engine's math is re-derived independently by its own test suite before a single question ships. Visuals pass a separate review so a diagram never accidentally reveals the answer.
When your child gets stuck, they can call the AI Tutor. It's a Claude-backed tutor trained to teach the way good teachers do: ask, listen, scaffold, redirect — and never just give the answer. Sessions are multi-turn, calibrated to your kid's mastery level, and bounded (capped at 5 teaching sessions per day, so no rabbit holes).
Any chatbot will happily do your child's homework for them. That's the problem. Our AI Tutor is built the other way around: it knows the exact question on your child's screen, it knows their mastery history — and it is mechanically prevented from just handing over the answer. Kids try. (Ours did. It's why we built it this way.)
| Algo School AI Tutor | ChatGPT & general chatbots | |
|---|---|---|
| Give away the answer? | Blocked by code, not just instructions — every reply is checked against the question's answer key before your child sees it. Paste the question, beg, retry: it coaches instead. | Will solve it outright on the first ask — that's what it's for. |
| Knows what your child is working on | Sees the live question, the concept, the grade level, and what your child has mastered or struggled with before. Picks up mid-lesson exactly where they left off — and remembers what happened last session. | Starts blank every time. Your child has to explain the problem — or paste it in. |
| Stays on the lesson | Scoped to the concept being learned. Off-topic asks get a friendly redirect back to the math — no drifting into games, videos, or "write me a story." | Follows the child anywhere they steer it. Distraction is one message away. |
| Shows, not just tells | Teaches with rendered diagrams — area models, number lines, fraction bars — the same visual language as the practice questions, drawn correctly by our own renderers. | Describes pictures in prose, or generates images that may get the math wrong. |
| Built-in limits | Bounded sessions (5 per day), independent safety screening on conversations, and a kid-safe kiosk with no browsing and no open internet. | Open-ended by design. Time limits and content boundaries are on you. |
| What you see as a parent | A brief after every session — the concept, what clicked, what stuck, and a dinner-table question to reinforce it. | Nothing. The conversation is invisible to you. |
Every session starts from a lesson plan built for your child: the tutor knows which step they're on, which misconception usually trips kids up here, and what your child answered in practice. It teaches in small moves — a question, a nudge, a quick check — and adjusts when the answer shows a misunderstanding. Close the tab mid-lesson? It saves the spot and picks the conversation back up, even days later, with a quick "welcome back" recap.
We're not the only option, and we're honest about that. Here's an even-handed look at where the major players shine and where we differ.
| Algo School | IXL | Khan Academy | iReady | Kumon | |
|---|---|---|---|---|---|
| Mastery model | Bloom-tagged, threshold-based | SmartScore (proprietary) | Skill tree progress | Diagnostic-driven path | Worksheet sequence + tests |
| AI tutor (Socratic) | Yes — AI Tutor, multi-turn | No | Khanmigo (paid add-on) | No | Instructor at center |
| Per-session parent brief | AI-written, emailed | Weekly summary | Limited | School-only reports | Biweekly center meeting |
| Question variety | 8 answer types, 50 visual types | High volume, repetitive | Video-heavy + practice | Diagnostic + practice | Paper worksheets only |
| Designed for | Parent-driven, at home | Schools + home subscriptions | Self-paced learners | Schools (district-licensed) | Center-based + home worksheets |
| Subjects ready today | Math G1 – Algebra 2 | All K-12 subjects | K-12 + test prep | Reading + math K-8 | Math + reading, K-12 |
| Free trial | 15 days, full access | Limited free preview | Free tier (with ads / limits) | School-licensed only | Varies by center |
If your school already provides iReady, that's your baseline. If you want broad K-12 coverage today, Khan or IXL are excellent. If a weekly in-person center fits your routine, Kumon is the classic. Algo School is for parents who want a clear plan, an AI tutor that actually teaches, and full visibility into what their kid did — without leaving the house.
You pick the concepts (or start from your child's grade), Algo School generates short daily practice sessions, checks every answer, and adapts the next session to what your child actually mastered. You see progress per concept — not a vague "time spent" number.
Short and daily beats long and rare. Sessions are sized to a time budget you set (most families use 10–30 minutes); the session builder fits the question count to your child's pace, not the other way around.
No — and not because we asked it nicely. Every tutor reply is checked against the question's answer key before your child sees it; if a reply would reveal the answer, it's blocked and rewritten as coaching. Kids trying to trick it (ours did) is a permanent test suite: if a trick ever works, our release fails before it ships. Tutoring is also capped per day and scoped to the concept being learned.
The practice library is built by 177 deterministic question engines we wrote — a fraction engine knows how to make a correct fraction question every time — and verified before it reaches your child. No hallucinated math, no wrong answer keys.
"Mastered" in Algo School requires evidence: at least 10 correct questions across at least two separate practice sessions, plus a later confirmation. One good streak in a single sitting doesn't count — which is exactly how memorization sneaks past other apps.
Start them on their grade's concepts; within a few sessions the mastery profile shows which concepts hold and which don't — per concept, with the questions to prove it. That's a diagnosis you can act on, not a percentile.
A dashboard with each child's daily status, progress charts, and what needs you — plus a short email brief after tutoring sessions with a question to ask at dinner.
$15 per child per month, all subjects included, after a 15-day full-access free trial. Progress data is never paywalled.
Algo School is built for every school subject. Math is live today — Grade 1 through Algebra 2 including Pre-Algebra, aligned to Common Core. High-school Geometry and Statistics are still being built and are not open for enrolment yet; ELA is next. We only claim what's shipped: your child can practice Grade 1 to Algebra 2 math right now.
No. You choose what to practice; the explanations, examples, and tutoring are built in. The dinner-table questions come with the answers.
Very well — homeschool IS the parent-driven model Algo School is built around. You set the plan from Common Core-aligned concepts (Grade 1 through Algebra 2), your child practices to evidenced mastery, and you keep a per-concept record of what's proven — useful for your own assessment rhythm. Explanations, worked examples, and the answer-guarded tutor are built in.
It's the practice-and-mastery spine, not a boxed curriculum: every concept comes with explanations, examples, and tutoring, but Algo School is built around practicing to proven mastery. Many homeschool families pair it with their preferred lesson source and let Algo School be the evidence that the math actually stuck.
That is the part Algo School is actually built around. Each concept carries the specific ways children go wrong on it — the mistaken rule, why it feels right to a child, and the sentence you are likely to hear them say. When your child practices, the concepts they are stuck on show you that diagnosis and the words to say back, and the tutor recognizes the pattern while they are working. You can read the public write-ups before you sign up.
Yes. Questions with symbolic answers get a built-in math pad: fraction bars, exponents, roots, inequality signs and Greek letters, with a live preview of what they have typed. The pad only shows the symbol groups that question can need, so a Grade 3 child never sees an integral sign. There is also a freehand scratchpad on the practice screen for working it out by hand.
Four kinds, and you can switch any of the first three off. A short brief after each AI Tutor session; a daily evening note when your child masters something; a Sunday-evening weekly digest of the week's practice; and — not opt-out-able, on purpose — an immediate alert if our safety screening ever flags something serious in a tutoring conversation.
Three reasons. Trust: a chatbot can't tell you whether the questions it just made are mathematically correct, age-right, or answer-leak-free — ours come from 177 deterministic engines with every answer verified. A plan: ChatGPT hands you a worksheet; Algo School runs a sequenced, adaptive queue all year and keeps score. Visibility: you see per-concept progress and what's coming next — a chatbot shows you nothing.
You set the rhythm during setup: 3, 5, or 7 practice days per week, between 10 and 45 minutes per day. Each practice day serves one session from your child's plan. Most families settle on 3-5 days at 30 minutes — steady beats long.
Yes — one family workspace, one sign-in for you, and a simple name-and-PIN kiosk for each kid. Each child gets their own grade, plan, and mastery profile. Billing is per child, so you only pay for kids actively practicing.
AI can be wrong — we say so plainly. The tutor is designed to ask questions rather than declare answers, so the usual failure mode is a clumsy question, not a wrong fact — and the practice questions themselves are engine-verified, never invented by the tutor. Every parent brief includes a one-tap report; we read every one.
Kids use aliases only — we never collect a child's real name. We don't sell data, ever, and card details live with Stripe, not on our servers.
Every time your kid practices, you get clearer signal — not a vague "time spent" number. Open the app and see the week at a glance, then drill into any concept down to the question.
We don't teach lessons. We teach concepts.
Free for 15 days — a card is kept on file, but $0 is charged until the trial ends. Cancel in one click before then and pay nothing.
Start your free 15 days →Math G1–Algebra 2 today, ELA next.