Every tool in the suite rests on the same idea about how learning is best structured. This page explains where that idea came from, what it assumes, and where it stops.
When several classes teach the same unit, they rarely move at the same pace. Some reach new ideas quickly; others need longer to consolidate. Over a term, the classes pull apart. We call this drift, and it shows up in two ways.
The first is pace drift. Classes reach the same ideas at different times, and without a plan for that, teachers start making informal adjustments that accumulate. Faster classes get moved on to new curriculum content. Slower classes fall behind because the dependencies they need are not yet secure, and without a way to address that, the gap compounds. Common assessment becomes harder to justify, and the shared reference point that planning depends on disappears.
The second is dependency drift. This is drift away from the principled pathway itself, and it can happen even without pacing pressure. A teacher follows what feels like a natural enrichment into a related idea, but that idea was not required for the unit's goal, and it rests on knowledge the students have not been taught yet. The detour costs time, and students end up building on ground that has not been laid. The instinct to enrich is a good one. Without a clear map of what depends on what, it just tends to lead away from the path rather than enriching the ideas already on it.
The suite was built to solve both at once.
Every unit has a sequence of ideas students must pass through to reach the curriculum goal. We call these waypoints. The order is not arbitrary. It follows the genuine dependencies in the content, so that each step rests on what needs to be in place before it. What carries that dependency differs by subject. Sometimes it is conceptual prerequisites, sometimes how a skill builds, sometimes the phases of making something, sometimes threads that progressively weave together. The commitment is the same in every case. Getting the order right means students build on solid ground instead of carrying gaps forward.
At each waypoint, classes may move at different speeds, and that is expected. Rather than advancing faster classes to new concepts, the model holds every class at the same waypoint and varies how broadly the concept is explored. We call this width. Some classes consolidate. Others widen their understanding through additional contexts, representations, and connections.
Every class moves through the same ideas and reaches the same goal. The shared sequence prevents dependency drift, because the map of dependencies makes clear what belongs on the path and what does not. Width absorbs pace drift without acceleration. And width gives the enrichment instinct somewhere legitimate to go. Faster progress leads broader into the same idea, not sideways into content the sequence has not prepared students for.
Width is easiest to see side by side. At the same waypoint, one class might work with a single representation and a familiar example. Another applies the same idea across several contexts, compares representations, or examines the misconceptions that commonly cluster there. Both classes are learning the same concept. The difference is how much cognitive demand is built in while they are there, not how far along they are.
Width means more contexts for the same concept, more connections to related ideas, more ways of representing it, and more chances to apply it somewhere new. It does not mean harder questions about the same example (that is depth), content from next year (that is acceleration), or reduced expectations for some classes (that is deficit thinking). Width is the dimension along which a shared waypoint flexes to meet a class where it is.
Widening takes different forms depending on what a waypoint calls for. Sometimes it means building fluency, sometimes varying contexts, sometimes accumulating approaches, sometimes making reasoning visible, and sometimes bringing parts together into a whole. The model distinguishes these, the tools draw on the distinctions when they generate plans and resources, and the distinctions themselves track established research on varied practice, schema formation, worked examples, and transfer.
The model is deliberately narrow. It is not a way to:
Every class remains on the same shared path at all times. That constraint is the point, not a limitation of it.
That is the idea. What follows is how it holds up in practice.
This is the shared vocabulary. These terms recur across every tool in the suite and form the common language that lets work in one tool carry into another.
Waypoints are not lesson activities, and they are not generated at random. They represent the dependency steps students move through to get from prior learning to the curriculum goal. Sequences are built from curriculum content descriptions and elaborations, known dependencies, and the breakdown points that commonly affect later learning.
Two habits keep the sequences honest. The starting point is traced, not assumed. Prior learning is followed back through the curriculum to establish what students genuinely arrive with, because a wrong starting point either wastes time on what is already known or leaves invisible gaps that surface later. And not every position in a sequence is a hard requirement. Some steps must sit where they do because later work genuinely depends on them. Others are placed by choice, because the position serves the unit well. The model distinguishes the two, so teachers know which parts of the order are fixed and which are theirs to move.
This approach fits subjects and units where learning can be organised as a shared progression toward a common goal, whatever form the dependencies take. It began in conceptually cumulative subjects like science and mathematics, and has since been extended and tested in units organised around skill development, design cycles, studied texts, and units that bundle several outcomes together. Those forms are now part of the model.
What remains out of scope is learning with no shared path at all: fully individual, emergent, or open-ended contexts where each student's progression is genuinely their own. The model is also honest about its own status. Parts of it are settled and have held up across subjects and year levels. Parts are working positions under active testing, and the open questions are treated as open rather than papered over.
Start with a real class you teach. Judge pace and consolidation needs from classroom evidence rather than assumptions. Adjust width as the unit unfolds, while keeping the sequence fixed. The goal is alignment, not uniformity. Classes may take journeys through a unit that feel quite different, and still arrive at the same understanding.
The same convictions that shape the model shape how the tools are built. Rather than being values statements, each one is a design decision you can verify by using the tools.
Your judgment is the mechanism, not a bottleneck. The tools don't call AI services directly. They build you a carefully structured prompt; you run it through whichever AI your school already trusts, and paste the result back. The Planner, for example, walks through several stages of review before generating anything, so you check each layer before the next one builds on it. That is slower than a button, and deliberately so. Every output passes through your hands before it touches a classroom, and outputs are written as starting points that expect editing, not scripts that expect compliance.
Nothing about you or your students is collected. Because generation happens in your AI rather than ours, your content never touches our servers. There are no accounts and no stored lesson data. Students join Signal sessions with a class code, and no student logins exist anywhere in the suite.
Productive struggle is protected. Topic Explorer's quizzes and matching activities let students retry but never reveal the answer. Its coaching prompts constrain the AI to guide within boundaries you set rather than solve on demand. Signal asks students how confident they feel, not whether they got it right. Insight classifies student thinking against waypoints without scoring it. These are signals for your decisions, not scores for theirs.
Your work stays yours. Decks are portable Markdown you can export to PowerPoint. Audio exports as ordinary MP3s. Plans export as files you can open anywhere. If the suite disappeared tomorrow, nothing you made with it would.
The ideas here are abstract until you watch them work. Start with a tool → · Browse the full suite →