2016 · CLASSROOM EVALUATION
Learning in real classrooms
A study of over 1,700 K–12 maths sessions, with more than 10,000 students using Adaptemy over six months.
Learning Effectiveness of Adaptive Learning in Real World Context
View publicationAdaptemy brings learning science into your digital products. Our engine builds a continuously updated understanding of each learner and orchestrates diagnosis, guided learning, practice and revision—using your content, inside your platform.
You define the learning goals and teaching approach. The engine adapts the learning direction, activities, support and challenge as each learner progresses. You keep your brand and customer relationships.

Our partners include educational publishers, learning platforms and national education providers. They bring the content and teaching expertise; Adaptemy supplies the adaptive intelligence.
Every learning interaction becomes evidence. Three connected models turn that evidence into an understanding of the learner and decisions about their learning experience.
Maps what learners need to understand and how concepts connect—including the prerequisites that underpin progress.
Connects your questions and learning resources to curriculum concepts and represents the characteristics needed to select suitable activities.
Continuously estimates knowledge, skills and retention, alongside uncertainty in those estimates and broader characteristics relevant to personalisation.
Together, these models determine where learning should focus, which learning strategy is appropriate and how the experience should adapt. Learner knowledge is modelled separately from individual resources, so your content can evolve while preserving the learner’s history.
The engine identifies the curriculum focus and an appropriate strategy, such as diagnostic assessment, guided learning, revision or exam preparation.
Within that strategy, it selects and sequences activities, adjusting support and challenge as the learner’s performance develops.
Knowledge cannot be observed directly. Adaptemy infers it from learning and assessment interactions, continuously updating its estimates as evidence accumulates.
Distinguishes longer-term competence from what a learner can demonstrate today, giving a more nuanced basis for learning decisions.
Represents uncertainty explicitly, helping distinguish a likely knowledge gap from an area where further assessment is needed.
Models how knowledge persists and recall changes, informing when learning should be revisited.
Can represent curriculum concepts, reasoning skills used across topics and finer-grained knowledge such as vocabulary.
The learner model combines established methods, including Item Response Theory and probabilistic inference across curriculum relationships, with Adaptemy’s proprietary modelling extensions.
An extended learner profile can also incorporate goals, motivation, self-efficacy and other characteristics relevant to your learning design. These complement the knowledge model and inform how the experience is personalised.
Your content is already an asset. Personalised learning creates new ways to package it, demonstrate its value and build a sustainable digital business.
Create a distinctive digital offer around diagnostic assessment, adaptive practice or revision. Give schools a clear learning benefit and your sales team a stronger value proposition.
Show teachers where learners are progressing and where support is needed. Give schools evidence to assess the value of continuing with your product.
Use suitable content across placement, mastery and exam preparation. Extend successful learning designs into additional subjects and age groups.
See how questions and learning resources perform in use. Prioritise what to improve, replace or commission with evidence from real learners.
We start by agreeing the business case with you: the product opportunity, the audience, the route to revenue and the measures that will tell us whether it is working.
Imagine a class practising percentages. The learning goal is shared; the help each student needs is different.
A learner struggling with percentages may need to revisit fractions. The engine uses evidence across related concepts to identify the likely gap and guide suitable practice.
A learner demonstrating understanding can progress to a more demanding activity. Challenge is informed by their learning history and readiness.
Your product can surface mastery estimates and likely gaps across the class, helping the teacher decide where to intervene and what to revisit together.
You define the goals and learning strategies. The engine uses the connected models to choose the learning focus and adapt activities within it—balancing challenge and support as proficiency develops. The same foundation supports diagnosis, mastery, intervention and revision.
Adaptemy’s engine and underlying approaches have been evaluated across educational settings, with published work spanning learning effectiveness, learning flow, sub-skill modelling and curriculum relationships. This research informs how we estimate learner knowledge and translate it into pedagogical decisions.
2016 · CLASSROOM EVALUATION
A study of over 1,700 K–12 maths sessions, with more than 10,000 students using Adaptemy over six months.
Learning Effectiveness of Adaptive Learning in Real World Context
View publication2021 · LEARNER MODELLING
Research into finer-grained learner evidence and targeted scaffolding to support progress.
A Conceptual Framework for Extending Domain Model of AI-enabled Adaptive Learning with Sub-skills Modelling
View publication2026 · CURRICULUM INTELLIGENCE
Research into using learner evidence to discover connections between concepts and explain curriculum structure.
Using Deep Knowledge Tracing to Discover Knowledge Concept Relations and an Explainable Curriculum Structure
Explore researchWe assess predictive performance, changes in estimated mastery, completion and learner satisfaction separately. These measures help you evaluate the product you bring to market.
A reported measure of how well the model distinguishes likely correct responses from incorrect ones.
A reported increase in estimated mastery during an adaptive session—not a measure of content completed.
Of 2,691 learners surveyed, 92% strongly agreed that they enjoyed adaptive learning more than a traditional lesson.
These figures are reported by Adaptemy across learning settings and describe different measures. Evaluation for your product is agreed against its learners, goals and context.
A project brings your publishing expertise together with our learning science and implementation team. We agree the scope, responsibilities and success measures before the build.
Agree the audience, subject, product proposition and business case. Choose a focused pilot or a wider implementation and define what success will look like.
Your learning designers explain your approach. We draft curriculum maps, support content tagging and configure learning strategies; your editors review and approve.
Your developers connect learning events and recommendations through the API. Your product team decides how the experience and insights appear to students and teachers.
Review learning and product performance against the agreed measures. Refine the experience and use the evidence to decide what to improve or extend next.
Your content stays in your systems. You retain your brand, pedagogy, interface and customer relationships. The engine works with learner identifiers without requiring names or email addresses.
A fixed professional-services fee covers the agreed implementation scope. An enterprise licence is charged per learner, per year. Pricing reflects your products, scale and selected capabilities.
We scope the schedule around your content readiness, learning design and integration. Your team contributes editorial review, learning-design decisions and development time. We agree milestones and responsibilities before work begins.
Discuss your project at FrankfurtMeet us in Frankfurt to explore a product opportunity, see how the learner, curriculum and content models work together, and discuss how your teaching approach could translate into an adaptive learning experience.