JabbaCalm on the outside. Real learning science inside.
Jabba is warm, slow and friendly on purpose. Every one of those choices rests on decades of research into how children actually learn. Here's the thinking underneath, in plain language.
Jabba is a learning companion, not a diagnosis or a treatment.
Ask about the England pilot
The design that includes is the design that works for everyone
One idea runs through everything Jabba does: the choices that make learning reachable for a child with ADHD, dyslexia or dyscalculia are the same ones that help every child learn more comfortably. Fewer things on the screen. More than one way to take information in. One small step at a time, and mistakes that never sting. That's the principle behind Universal Design for Learning, the inclusive-education framework schools use worldwide. Here, the different ways to see, hear and respond are on by default, not bolted on as 'special settings'.
First the foundation, then the next step
Mastery learning only works when you know what builds on what. We mapped the National Curriculum in England, decomposed its assessable statements into teachable learning goals, and independently reviewed the prerequisite links. This is the actual England graph used by the test rollout. Explore it below — rotate, search, and select a dot to follow the chain.
- 12National Curriculum subjects across Key Stages 1–4
- 1,495assessable curriculum statements mapped to their official source
- 3,947reviewed learning goals — small enough to practise one at a time
- 4,102reviewed prerequisite links — what builds towards the next capability
Loading…
The map shown here is England graph v14, approved for the controlled task pilot.
One-to-one tutoring is one of the most powerful tools in education. AI finally puts it within reach.
For forty years the research has said the same thing: a child working with a patient one-to-one tutor learns far more than the same child in a busy classroom. It's one of the largest effects we've ever measured in education. The catch has always been cost — no school can give every child their own tutor. That's the gap a well-designed AI tutor can close, and the first careful studies are now showing it works.
Among the largest effects in education
One-to-one mastery tutoring has outperformed the ordinary classroom by a wide margin ever since Bloom's landmark 1984 study.*
Twice the learning, in less time
A 2025 Harvard randomised trial found students with a well-designed AI tutor learned more than twice as much as in an active-learning class — and finished sooner.*
On par with a human tutor
A meta-analysis of 14,000+ learners found AI tutoring systems matched one-to-one human tutoring and beat whole-class teaching.*
The gains come from the teaching, not the technology. Those studies worked because the AI was carefully designed to teach well — short replies, one step at a time, hints instead of answers. That's exactly how Jabba is built. A chatbot that simply hands over the answer does the opposite.
* A finding about the teaching method, drawn from the wider research, not a measurement of children using Jabba.
See the science show up in the app
Swipe through the features that put the research to work, and see how each one looks in Jabba.
* A finding about the teaching method, drawn from the wider research, not a measurement of children using Jabba.
Always the right level of challenge
Learning happens fastest in the gap between what a child can do alone and what they can do with a little help. Jabba lives in that gap. It reads how each step goes and re-pitches the next one in real time: more support when a child struggles, a stretch when they're flying. It never just hands over the answer, though. Your child always writes that part themselves, and when something's wrong, Jabba gently explains why.
Tutoring that keeps a child in this 'sweet spot' is linked in research to roughly a third more progress in maths.*
Numbers your child can see and move
For maths, Jabba starts with things a child can actually see and move: blocks, number lines, fraction bars, shown right beside the symbols they stand for. Then it bridges to the numbers alone. Research links this concrete-to-abstract approach to stronger maths learning, especially for children who find numbers hard, and it builds a firmer foundation for everyone.
Reading with all the senses
Letters and sounds are practised by seeing them, hearing them and saying them. Every word can be read aloud too, so trouble with decoding never blocks a child from the meaning. It draws on the explicit, multisensory approach research backs for dyslexic readers, with a dyslexia-friendly font and one chunk of text at a time rather than a wall of words.
Calm is how attention works
Working memory is small and easily crowded, and everything extra on a screen competes for it. So Jabba keeps things quiet: calm colours, no flashing, no ads, one thing at a time, and a visible 'you are here' tracker. A quieter screen means fewer distractions, so more of a child's attention stays where it should. For a child with working-memory differences, that's the difference between reachable and overwhelming.
Review timed to make it stick
Two of the best-replicated findings in learning science: recalling something beats re-reading it, and review spread over time beats cramming. Jabba notices how well each idea landed and brings it back at the right moment. Soon if it was shaky, weeks later once it's secure. It feels like a quick game, never a test.
Combining spaced practice with retrieval improves remembering by around a quarter over either alone.*
Safety comes first, by design
When a child feels anxious or ashamed, the brain quietly shuts the learning parts down. When they feel safe, it opens them back up. So Jabba is built so a child can never 'get it wrong' in a way that hurts: no grades, no red marks, nothing to break by missing a day. And if Jabba notices a child is really struggling, it pauses and points gently to a grown-up and the children's helpline. A child learns best from a place of safety, so that's where Jabba starts.
* A finding about the teaching method, drawn from the wider research, not a measurement of children using Jabba.
The National Curriculum, translated into the next teachable step
The National Curriculum sets out what pupils in England should learn across Key Stages 1–4. Jabba turns that large statutory framework into smaller, assessable learning goals and follows the reviewed prerequisite links between them. The aim is simple: home education and supplementary learning should reinforce the same foundations instead of becoming a disconnected collection of exercises.
Help with confidence, no teaching degree required
You don't need to know the pedagogy to use it. The parent overview shows you the day's activities, how your child is doing and how they're feeling, always framed as growth rather than deficit. And a try-it-yourself sandbox ("Prøv selv") lets you try an activity yourself before your child does.
A note on honesty
Jabba is an educational companion: not a diagnosis, not a treatment, and not a substitute for specialist help. It's designed around how children who learn differently learn, which makes it better for every child, but it does no screening or therapy. Where we cite research, we're describing the methods Jabba is built on, not promising a specific result for your child.
Some of the research this rests on
- Universal Design for Learning — CAST (2024), UDL Guidelines v3.0; Capp (2017), International Journal of Inclusive Education, meta-analysis.
- Zone of proximal development & scaffolding — Vygotsky (1978), Mind in Society (Harvard University Press); Wood, Bruner & Ross (1976), Journal of Child Psychology and Psychiatry; van de Pol et al. (2010), Educational Psychology Review.
- Concrete-to-abstract maths — Butterworth, Varma & Laurillard (2011), Science; Gersten et al. (2009), Review of Educational Research, meta-analysis.
- Structured, multisensory literacy — International Dyslexia Association (2019), Structured Literacy; Galuschka et al. (2014), PLOS ONE, meta-analysis of RCTs.
- Cognitive load theory — Sweller (1988), Cognitive Science; Mayer (2009), Multimedia Learning (Cambridge University Press).
- Retrieval practice & spacing — Roediger & Karpicke (2006), Psychological Science; Pan & Rickard (2018), Psychological Bulletin, meta-analysis.
- Emotional safety & SEL — CASEL (2020), SEL framework; Immordino-Yang & Damasio (2007), Mind, Brain, and Education.
- One-to-one tutoring & mastery learning — Bloom (1984), Educational Researcher; Kraft (2020), Educational Researcher, on effect-size expectations.
- AI & intelligent tutoring systems — Ma et al. (2014), Journal of Educational Psychology, meta-analysis; Kulik & Fletcher (2016), Review of Educational Research; Kestin et al. (2025), Scientific Reports, randomised controlled trial.
- National Curriculum in England — Department for Education programmes of study for Key Stages 1–4; source material used under the Open Government Licence v3.0.
What parents ask most
What learning science is Jabba built on?
Jabba rests on well-evidenced pedagogy: Universal Design for Learning, scaffolding within the zone of proximal development, concrete-to-abstract maths, structured multisensory reading, retrieval practice with spaced review, and cognitive load theory. The same principles inclusive schools use are switched on here by default rather than bolted on as special settings.
Is Jabba suitable for children with ADHD, dyslexia or dyscalculia?
Yes. Jabba is inclusive by design: fewer things on screen, more ways to see, hear and respond, and one small step at a time. That makes learning more reachable for children with ADHD, dyslexia or dyscalculia, and calmer for every child. Jabba is a learning companion, not a diagnosis and not a treatment.
Does Jabba follow the National Curriculum in England?
The England pilot is built around the National Curriculum from Key Stage 1 to Key Stage 4. Each approved activity is linked to a reviewed curriculum goal and its source, so the plan can follow what a child is ready to learn next. The task corpus is still being tested before public England signup opens.
Does Jabba just give my child the answer?
No. Jabba never hands over the answer itself. It re-pitches the challenge in real time and adds support when something is hard, but your child always writes the answer. When something is wrong, Jabba gently explains why, so the child understands it rather than just having it corrected.
Does Jabba replace a teacher or a specialist?
No. Jabba is a supplementary learning tool, not a replacement for a teacher, an assessment or specialist help. It does no screening or therapy. It is designed around how children who learn differently learn best, which makes it better for every child.
Is an AI tutor actually a good way to learn?
The evidence is encouraging. One-to-one tutoring is one of the most effective things in all of education, and recent randomised trials — including a 2025 Harvard study — found that a well-designed AI tutor can match or beat classroom teaching, with children learning more in less time. The gains come from good teaching, not the technology: Jabba is built to give hints rather than answers, one small step at a time, which is exactly what made those studies work.
See what England's curriculum looks like as a learning map.
Explore the reviewed graph above, then contact us if you would like to help test the first British-English task pilot.
Ask about the England pilot