You choose the concepts. Algo School generates the daily practice from 99,079 verified questions across 2,232 Common Core concepts, grades every answer, proves what your child has actually mastered, and hands you the coaching for what they get wrong. Grade 1 through Algebra 2.
Most school software assumes a teacher on one side and a parent receiving reports on the other. In your house those are the same person, which is why so much of it fits badly: you get a dashboard designed to inform someone who is not making the teaching decisions, and a curriculum designed to make them for you.
Algo School starts from the opposite assumption. The adult choosing what gets studied, the adult who will sit down and explain it, and the adult reading the evidence afterwards are one person — so the plan is yours to set, the practice runs without you, and the part that usually needs a teacher's training (working out what the child actually misunderstands, and what to say about it) is written out for you in plain language. You are not being kept informed. You are being equipped.
Algo School can be the spine of your core math program: you plan the year from the concept map, your child practices to evidenced mastery, and you get the diagnosis and the coaching script for every concept that wobbles. What it does not do is deliver the first lesson — it is not a boxed curriculum with a teacher's manual and a scripted introduction to each topic. You supply the instruction, from your own explanation or whatever lesson source you like; Algo School carries the practice, the grading, the mastery evidence, and the teaching support that comes after it. Families running it as their core are, in effect, using it as everything except the lesson.
It also runs alongside any curriculum you already own — Saxon, Math-U-See, Khan Academy, a co-op syllabus, a textbook from a used-book sale. Nothing has to line up, because you pick the band and the domains yourself instead of inheriting someone else's sequence. If your lesson this week is long division, you practice long division; the rest of the map waits. That is the same freedom the core users get, exercised in smaller pieces.
Everything starts from the knowledge graph — 2,232 Common Core concepts, arranged under their real domains, from Grade 1 through Algebra 2. You enrol a child in a band and then choose which domains they work through, and the system sequences the concepts inside them for you, backfilling prerequisites so nothing arrives before the thing it depends on.
There is no grade-lock, and that is the part homeschoolers care about most. When you set a child up, every band is offered — grouped as suggested for their grade, higher grades, and lower grades — so a nine-year-old can sit in Grade 5 fractions and Grade 3 measurement in the same week, and a fourteen-year-old can go back and rebuild place value without anyone treating it as remedial. The content follows the band you picked, never the age on the account. Pacing freedom is the whole reason you homeschool; it would be strange for the practice software to take it back.
You also choose HOW the plan is built: work through a band in academic-year order, target the concepts your child's own results say are weak, or hand-pick the topics yourself.
The two jobs that quietly eat a homeschool parent's evening are producing the practice and marking it. Both are gone. The library holds 99,079 verified questions, and they are not improvised by a chatbot on request: 177 deterministic question engines generate them — a fractions engine knows how to build a correct fractions question every time — and each bank is verified before a child ever sees it. Wrong answer keys are the one defect a practice tool cannot afford, so that is the defect the whole pipeline is built to prevent.
Every answer is graded the moment it is submitted, which means your child finds out while they still remember what they were thinking, and you are never handed a stack of paper at 9pm. Each session is drawn fresh from the plan, so repeating a concept does not mean repeating the same twenty questions.
"We covered it" is the weakest sentence in home education, and every homeschooling parent has said it about something their child could not do three weeks later. Algo School will not say it. A concept is marked mastered only when the evidence is there: at least two separate practice sessions on it, with accuracy holding up over the recent questions as well as over the longer stretch. One strong afternoon does not qualify — which is exactly how memorising sneaks past a checklist.
The judgement is re-made as your child keeps working, so a concept that quietly decays shows up as needing review rather than staying ticked off forever. What you get is a per-concept picture of what is proven, what is in progress, and what has slipped — and the sessions behind each verdict, so you can go and look rather than take our word for it.
A trained teacher's real advantage is not knowing more math than you. It is having seen two hundred children make the same specific mistake, so that when a child says "eight is bigger than four, so one-eighth is bigger than one-fourth", the teacher recognises it instantly, knows why it feels true to the child, and has a move ready. That recognition is what a teacher's guide sells you, and it is the thing a homeschooling parent most often does not have.
Algo School ships it as a feature. When your child struggles with a concept, you do not get a percentage — you get a diagnosis in parent language. What we think is going on, in one sentence. What you might hear, quoted in the words a child that age actually uses, which is usually the moment a parent recognises their own kid. Why it happens, because a mistake you understand is one you can address instead of just correcting. And how to help: the specific move that works on that specific misunderstanding.
Behind that sits the full coaching guide for the standard: a short teach-tonight pass you can run at the kitchen table, the question to ask and what a good answer looks like, the exact words to open with, the phrasing to avoid, the other mistakes commonly seen on that topic, and quick checks to run before sending your child back to practice. It is written for the adult doing the teaching — because in your house, that is who is reading it.
None of this is a demo you have to sign up to see. A slice of the same misconception library is published openly, one page per concept — the mistake, what it sounds like, and why children make it:
You have other children. Some days you are teaching one of them phonics while another is three questions into decimals and has gone quiet. A homeschool day is decided by what happens in those minutes, so the practice session carries its own help, six tools deep, in escalating order — a hint on the current question, an explanation of the concept, a worked example, a video, the AI tutor, and a scratchpad for working it out by hand.
The tutor is the one that deserves a caveat, so here it is plainly: it cannot give your child the answer. Not because it was asked politely in a prompt, but because every reply is checked against the question's answer key before your child sees it, and a reply that would reveal the answer is blocked and rewritten as coaching. Children trying to trick it — ours did — is a permanent test suite; if a trick ever works, the release fails before it ships. Tutoring is capped per day and scoped to the concept at hand, and you get a short brief afterwards saying what was discussed. Your child gets unstuck; nobody gets a homework machine.
Answers are typed in real notation, not squeezed into whatever a phone keyboard offers. An on-screen symbol pad puts π, ±, ≤, ≥, ≠, superscripts, subscripts and roots one tap away, and it shows only the groups the current question needs, with the rest behind a "show all symbols" control. A child answering an exponential-growth question types the exponent as an exponent.
This matters more at home than at school. Notation is normally absorbed from a whiteboard over years of watching a teacher write; a child practicing at the kitchen table sees far less of it. Typing it themselves, from Grade 1 arithmetic through Algebra 2, is how the habit forms.
Each child has their own dashboard: practice over the last seven days, concepts mastered, recent sessions, where they could use a hand, what is going well, and what is currently being learned. It is a per-concept record of work done and mastery earned, kept continuously rather than reconstructed at the end of a year from memory and a pile of worksheets.
On Sunday evening, in your own timezone, a weekly digest lands in your inbox summarising the week each child had. Nothing happened that week means nothing is sent — an empty email is worse than no email. And to be straight about the boundary: this is your own record-keeping, on your own screen. Algo School does not produce paperwork for an oversight body, and we would rather say so than let you find out in April.
Homeschools are rarely one child. One sign-in belongs to you; each child gets a name-and-PIN kiosk they can open themselves, without an email address and without a password you will end up resetting. Behind it, each has their own grade, plan, session length and mastery profile — siblings two years apart are not sharing a queue.
You set the rhythm per child: 3, 5 or 7 practice sessions a week, at 30 or 45 minutes each. Practice is a queue rather than a timetable — one session is handed out at a time and the next appears when it is finished, up to that weekly number — so a heavy co-op day or a sick week does not leave a trail of missed assignments to clear. Billing is per child, so a child taking a term off is not a child you are paying for.
Sunday, you plan. Ten minutes with the concept map: this child stays in the fractions domain, that one moves up a band, the third needs the measurement work you skipped in the autumn. Monday through Thursday, each child opens their kiosk and works their session — 30 minutes, hints and worked examples and the tutor there if they stall, you in the next room doing something else. Everything is graded as they go.
Midweek you glance at the dashboard and see one concept flagged. You read the diagnosis, recognise the sentence your daughter said at breakfast, open the coaching guide, and spend eight minutes on it after lunch with the words already written out. She goes back to practice; the quick checks tell you whether it landed. Sunday evening the digest arrives with the week in one page, and you sit down to plan the next one already knowing what stuck.
It can be the spine of it, and we would rather be precise than flattering: Algo School plans the concepts, generates the practice, grades every answer, tracks mastery, and hands you the coaching for what your child gets wrong. It does not deliver the first-time instruction — you do, or your lesson source does. Families who use it as their core math pair it with their own teaching and let Algo School carry everything after the lesson.
Yes, and nothing needs to line up. You choose the band and the domains you want practiced, so Algo School follows your sequence instead of imposing one. Most supplement users pick the two or three domains their current lesson is in and let the rest wait.
Yes — that is the normal case here, not an exception. When you enrol a child, every launched subject is offered, grouped as suggested for their grade, higher grades, and lower grades. A ten-year-old can sit in Grade 3 fractions and Pre-Algebra at the same time; the content follows the subject you picked, not the age on the account.
You set it: 3, 5, or 7 practice sessions a week, at 30 or 45 minutes each. There is no feed, no autoplay, no streak pressure to come back — a session ends and the screen is done. Everything off-screen stays off-screen: the coaching is written for you to say out loud, not for your child to watch.
That is the part Algo School was actually built for. For a concept your child is struggling with, you get the diagnosis in plain language, the sentence you are likely to hear them say, why the mistake makes sense to them, and what to say next — plus a coaching guide with a short teach-tonight pass, the words to use, the mistakes to watch for, and quick checks before they go back to practice. In the meantime your child is not stranded: hints, an explanation, a worked example, a video, a scratchpad, and an AI tutor that is blocked from handing over the answer are all one tap away.
Mastery here is earned, not awarded. A concept counts as mastered only after at least two separate practice sessions, with the accuracy holding up across the recent window as well as the longer one — so one strong afternoon does not qualify. Your dashboard shows it per concept, with the sessions behind each verdict.
Yes. One sign-in for you, one family workspace, and a name-and-PIN kiosk for each child. Every child has their own grade, their own plan, their own session length, and their own mastery profile. Billing is per child, so you pay only for the ones actively practicing.
Yes — Common Core, Grade 1 through Algebra 2, including Geometry and Statistics. Concepts sit under their real Common Core domains, so when you plan a week you are choosing from the same map a school would use. The library is 99,079 verified questions built by deterministic engines, not generated on the fly by a chatbot.
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 →