Adaptive practice
Question sets that respond to the individual student — easing off when something is secure, slowing down and re-teaching when it isn't. The path through a topic is different for every learner, because it should be.
UK education technology
Learning technology built for how students actually learn.
Bnm makes digital learning tools that adapt to the individual student — so progress is visible, teaching decisions are grounded in evidence, and no one quietly falls behind.
Built with this in mind
UK data residency
UK GDPR by design
WCAG 2.2 AA target
Evidence-led design
Mastery reached
Needs a nudge
92%
61%
34%
Quadratic factorising
Secure — moving on
Simultaneous equations
Developing — extra practice set
Rearranging formulae
Misconception flagged to teacher
Illustration of the teacher view. Sample data.
Why we exist
A classroom moves at one speed. Thirty students don't.
Every teacher knows the problem. Some students needed the lesson to slow down three examples ago; others were ready to move on before it started. By the time a gap shows up in an assessment, it has usually been widening for weeks.
Bnm exists to close that gap while it is still small — and to give teachers the evidence to act without adding hours to their week.
What we build
Three things, done properly.
We would rather do a small number of things well than ship a dashboard nobody opens twice.
Insight teachers use
Not a wall of charts. A short, honest answer to the question teachers actually ask on a Monday morning: who is stuck, on what, and what should I do about it in tomorrow's lesson?
Motivation that lasts
Progress students can see, framed around mastery rather than leaderboards. We are careful with streaks, badges and anything else that trades long-term confidence for short-term engagement numbers.
For students
The right next question, every time.
A student working through a topic on Bnm is never guessing what to do next. The platform reads what they have just demonstrated and picks accordingly — a harder variation, a worked example, or a step back to the sub-skill that is actually causing the trouble.
- Diagnostic, not just marking. A wrong answer tells us which misconception is behind it, not only that it was wrong.
- Explanations at the point of need. Support appears where the student is stuck, not buried in a separate resources tab.
- Spaced retrieval built in. Topics come back before they are forgotten, on a schedule set by the evidence on memory.
- Works on the device they have. Phone, Chromebook or ageing school desktop — including on a bad connection.
Student view
Solve for x: 3(x − 4) = 2x + 5
x = 7
x = 17
x = −17
Nice — that's secure.
You've got three of these right in a row. Moving you on to equations with brackets on both sides.
You've got three of these right in a row. Moving you on to equations with brackets on both sides.
For teachers
Evidence, not admin.
Most classroom software adds a task to a teacher's day. Ours is judged on whether it removes one. Marking is automatic, gaps are surfaced without being hunted for, and the class view is designed to be read in the two minutes before a lesson starts.
- Misconceptions grouped, not listed. See the four students making the same error, and teach it once.
- Set work in under a minute. Or let the platform assign follow-up practice from what it already knows.
- Exportable for reports. Progress data that goes into your existing reporting cycle instead of duplicating it.
- You stay in charge. Every automated suggestion can be overridden. The platform advises; the teacher decides.
Needs attention
Four students share the same misconception in Rearranging formulae:
Dividing only one term when
undoing a coefficient.
AK Amira K.
TO Tom O.
JW Jess W.
DR Dan R.
Illustration. Sample data.
How it works
Running in a fortnight, not a term.
Rolling out school software is usually the hard part. We have tried to make ours the easy part.
1
Baseline
A short diagnostic establishes where each student genuinely is — not where the scheme of work assumes they are. Typically one lesson.
2
Adapt
Students work through practice that responds to them. Teachers see gaps as they emerge and set targeted follow-up in a couple of clicks.
3
Review
Department and leadership views show movement over time, so you can tell what is working and where support should go next.
What we will not do
Some products are defined by what they refuse to ship.
We work with children's data and children's confidence. Both deserve more caution than the sector usually applies.
No selling student data. Ever.
Not to advertisers, not to data brokers, not to "analytics partners". It is not a revenue line we are choosing to forgo — it is one we will not build.
No engagement traps
No mechanics designed to make a fourteen-year-old anxious about breaking a streak. Motivation should come from getting better at something.
No black-box grading
If the platform says a student is struggling, a teacher can see exactly which responses led to that conclusion, and disagree with it.
No replacing teachers
We are building an instrument, not a substitute. Every design decision assumes a skilled adult is in the room and knows the student better than we do.
Who it's for
One platform, three points of view.
Students
Clear next steps, honest feedback and visible progress. Nothing that makes falling behind feel like a public event.
Teachers
Marking handled, misconceptions surfaced, and planning informed by what happened in the last lesson rather than guesswork.
School leaders
Department-level movement, intervention tracking and a defensible evidence trail — without another data-entry burden on staff.
Questions
The things people ask first.
Which subjects and key stages are covered?
Our first release focuses on secondary mathematics, mapped to the English national curriculum. Additional subjects follow once we are confident the underlying model holds up in a second domain — we would rather be genuinely good at one subject than shallow across eight. Talk to us about your specific requirements and timelines.
Where is student data stored, and who can see it?
Student data is held in the UK and processed under UK GDPR. Your school remains the data controller; we act as processor under a written agreement. We do not sell data, and we do not use identifiable student data to train models for other customers. Full detail is in our privacy notice.
Does this replace teacher judgement?
No, and it is designed not to. Every flag the platform raises is traceable to the specific responses behind it, and every automated suggestion can be overridden. The system's job is to notice things quickly across thirty students at once — the teaching decision stays with the teacher.
What does it cost?
Pricing is per student per year, with the rate depending on cohort size and the length of the agreement. We would rather quote you honestly against your actual numbers than publish a headline figure that turns out not to apply. Get in touch and we will send a written quote.
Will it work on our school's devices?
It runs in the browser, with no installation and no plugin. It is built to work on Chromebooks, iPads, older Windows desktops and student phones, and to degrade gracefully on a slow or intermittent connection — which, in a lot of schools, is the actual constraint.
How much work is the rollout for our staff?
We import your class lists, run a short session with the department, and stay close through the first half-term. The baseline diagnostic takes roughly one lesson per class. We have deliberately kept setup light because a rollout that needs three INSET days will not happen.
Early access
We are working with a small number of schools.
We are deliberately keeping our first cohort small so we can stay close to the people using it. If the problem described on this page is one you recognise, we would like to hear how it shows up in your setting.