Skip to content
thelearninglibrary.org
Our solution

Teachers should spend their time on students, not on building course material.

One shared library of study material for every textbook and subject, built once and kept fresh for every class. And a growth hub on top: what each student has mastered, what’s working for the class, and where each learner’s path leads.

The problem

Every teacher rebuilds the same worksheets, quizzes and tests on their own, for the same textbooks and the same standards. Those hours come out of evenings and out of time with students.

What we found

We reviewed the official assessment standards of 17 countries and provinces. They agree on what good assessment looks like: tasks tied to a published standard, results on a few described levels, practice kept separate from grading. Almost none of them supply the materials. That gap lands on teachers.

Read what the standards say

How it works

Three steps, from a textbook to a class that’s ready.

  1. Step 1: Find or add a shelf

    Search by title, ISBN or topic. If your book isn’t here, upload it once, up to 2,000 pages. Your file stays private.

    Browse shelves
  2. Step 2: Everything is built once

    A summary, a one-page cheat sheet, a full study guide and flashcards, written chapter by chapter from the book’s own pages and shared with every teacher who uses the same book.

    See a shelfHow a big book gets built
  3. Step 3: Fresh papers and a study plan

    A new worksheet, quiz or test in one click, every time. Students get a Learn tab that adapts to what works for them, one chapter at a time.

    Try the test generatorHow Learn adapts

How we use AI

AI does the reading and the first draft, so no teacher has to. Every score that counts (check-ins, mastery checks, the learning profile) and every check on the AI itself is plain code we can explain.

From a book to a shelf

  • No AIPlain code: checks, maths and scheduling we can explain line by line.
  • Fast modelClaude Haiku: quick, cheap judgements like naming chapters.
  • Writing modelClaude Sonnet: reads every chapter and writes the material.
  • Strongest modelClaude Opus: the two whole-book pieces, where judgement matters most.
  1. Step 1Size and safety checkPage count read in your browser before upload; malware scan; anything over 2,000 pages never leaves your computer.No AI
  2. Step 2ScreenSpam, ads and prompt injection are refused before any real work is spent on them.No AIFast model
  3. Step 3Find the chaptersBookmarks, contents and headings give candidates; a model gives each chapter its real name and sets aside the preface, index and answer keys.No AIFast model
  4. Step 4Read every chapterLong chapters are split at natural breaks and read in parallel. Out come the key concepts, each with the pages it’s taught on.Writing model
  5. Step 5Write the materialStudy guide from the concepts and the book’s own pages; flashcards from the concepts. Checked for copying and rewritten if needed.Writing modelFast model
  6. Step 6The whole bookSummary and cheat sheet from the most important ideas of every chapter.Strongest model
  7. Step 7Place it on the mapCompared with nearby books: what comes before, what comes after, what teaches the same thing.Fast model

What AI does

  • Reads each chapter of a book and pulls out its key ideas, with page references
  • Writes the summary, cheat sheet, study guide and flashcards in its own words
  • Writes fresh questions for every worksheet, quiz, test and mastery check
  • Gives feedback and asks guiding questions when a student explains an idea back or teaches it
  • Suggests how books connect: what to learn first, next, or instead

What AI never does

  • Scores a mastery check or a check-in: those are marked against an answer key
  • Decides a student’s plan, pace or review schedule: that’s plain maths (FSRS and our check-in rules)
  • Hands a student the answer while they practise
  • Writes about the book itself, like its publisher or edition; only the subject
  • Shows anyone the uploaded file
New

A map of what to learn next

Like a skill tree in a game, every book in the library links to its neighbours. Finish one and there’s a next step; get stuck and there’s another way in. The map builds itself as shelves are added, and teachers can set links by hand.

Algebra Ilearn firstAlgebra IIyou are herePrecalculuslearn nextStatisticslearn nextAlgebra IIexplained differentlybackforwardsideways
An example. Each link carries a one-line reason a student can read, and shows the version of a book with the most likes. How the map is made
  • Learn first

    Books that teach what this one assumes. When a chapter won’t click, Learn suggests brushing up on the groundwork.

  • Learn next

    Books that build on this one. Master every chapter and Learn shows you where to go next.

  • Explained differently

    Another book on the same material at the same level, for when one explanation isn’t working. Progress is kept.

How Learn works

One chapter at a time, round a loop that measures what sticks.

  • Each day opens with a short check-in on ideas from days ago, the honest measure of what was remembered.
  • Today’s plan mixes due reviews with new concepts, at a pace set by those check-ins.
  • Practice uses whichever study methods are working for this student, from 20 tools in 7 method types.
  • A mastery check of fresh questions opens the next chapter at 85%. Earlier chapters stay open for review.
Why chapter by chapterHow check-ins work
your currentchapterCheck-inwhat stuck from days agoToday’s planreviews + new conceptsPractice20 tools, 7 method typesMasterycheckfresh questionsNext chapterat 85% or moreLearning mapif it isn’t clickingtinted = AI involved
AI writes the practice feedback and the mastery-check questions, and it asks rather than answers. Check-ins, plans, review schedules and every score are plain code, so we always know what a student can do on their own.

Not just a library: a growth hub

The same data that adapts Learn shows students where they’re heading and shows teachers where to help, with nothing shared unless the student agrees.

  • Learn

    Master one chapter, then the next

    Check-ins find which study methods work for each student; mastery checks at 85% open the next chapter.

    Find a shelfRead more
  • Teach

    See who needs a hand, and what's working

    Each student's chapters, daily activity and latest check, plus which methods work across the class and which ideas most students miss.

    Make a classRead more
  • Grow

    Know where your learning leads

    An expertise web built from what you've mastered, compared with any job or college major, with the shelves that close the gap.

    Open My growthRead more
An illustration: a student’s mastered knowledge (green) against what a science job needs (dashed), on O*NET’s 1–7 knowledge levels. Real comparisons use O*NET’s data for the job you pick.

What we believe

Ten beliefs decide what we build, and what we leave out.

Aligned to your standards

Coming soon

Every question will say which official standard it checks, using the open CASE format, and results will read the same way your school reports them.

  • Common Core
  • Next Generation Science
  • Ontario
  • British Columbia
  • Australian Curriculum v9
  • New Zealand Curriculum
  • England National Curriculum
  • CEFR for languages

Our privacy promises

  • Uploaded textbooks stay private. Nobody can view or download them; only the study material is shared.
  • A student’s learning history is theirs alone, and they can delete it at any time.
  • We ask a parent’s permission before storing anything about younger students.
  • Teachers see a student’s progress only after the student agrees, and every view is logged.
  • We never sell data, show ads, or use student or teacher data to train AI.
Trust & student data

Get your evenings back.

Search for the textbook you teach from. If it isn’t here yet, add it and the shelf builds itself.