Innovating Math Education Through
AI-Powered Games
Age of Learning is an EdTech company offering research-based digital learning programs like ABCmouse, which have educated over 50 million children worldwide. In 2015, they began developing Math Mastery games with AI-driven, adaptive learning paths to accelerate skill mastery, enhance math fluency, and boost student confidence.
Role
Senior UX Design
Timeframe
10 months, 2015
Tasks
UX design, research and management
Industry
EdTech
Product Category
Consumer Product and Enterprise Tool
Challenges
Keeping kids motivated and challenged: ABCmouse.com’s fixed, linear game progression often failed to meet children at their skill level—either moving too fast or too slow—causing them to lose interest. This highlighted the need for adaptive, personalized learning experiences that could adjust in real time to each learner.
Staying ahead in a shifting market: In 2015, AI-powered learning was still in its early stages. Age of Learning recognized a strategic opportunity to lead the industry by embedding intelligent, adaptive technology into early childhood education products.
Objectives
Develop predictive AI math games integrating proprietary curriculum and technology to create an personalized, adaptive learning experience.
Leverage the Zone of Proximal Development (ZPD) pedagogical framework with machine learning to observe how students interact with the content, identify patterns and predicts what the student is ready for next.
Expand Age of Learning’s product offerings to reach older students, increasing engagement and market potential.
Personas
Provisional personas were created based on user insights and business goals to guide design decisions, ensure alignment with user needs, and serve as a reference for usability studies. These personas helped the team anticipate user behaviors, identify pain points, and validate design choices throughout the development process.
AI’s Role in Gameplay
Difficulty Adjustment: Questions get harder or easier based on performance.
Content Sequencing: The next topic or game level is chosen based on what the learner needs most.
Personalized Feedback: AI delivers tailored hints based on the student’s specific mistakes.
Peer-Informed Suggestions: Recommendations are guided by what’s worked for similar learners.
Educator-Driven: Teacher-defined goals and rubrics help shape content sequencing and focus.
Competitive Analysis
Competitive analysis of the key players in the industry was performed to identify market trends, uncover opportunities for differentiation, and inform strategic design and product decisions.
Testing Early with Real Learners
A usability testing practice was established for the Math Mastery games, involving weekly sessions to assess curriculum efficacy, refine game mechanics, and iterate on prototypes. Each session included moderated in-depth interviews and observations, with children accompanied by caregivers.
To create an optimal testing environment, we started by setting up a child-friendly testing room in the office designed to make participants feel comfortable and engaged while providing reliable user feedback.
Design principles
Design for Learner Agency and Flow – Give students choice in topics and paths. Use clear, encouraging feedback (“You’re doing well with decimals!”) and show progress with visual cues like progress bars.
Balance Challenge and Support (ZPD) – Use adaptive game mechanics like “Boost Mode” to adjust difficulty. Make hints and scaffolding easy to access, and add friction when students struggle to reinforce learning.
Design the Feedback Loop – Provide instant, motivating feedback with growth mindset cues (“Let’s try again—remember this part?”) to guide learners and encourage progress.
Teacher dashboard
In addition to the Math games, a Teacher tool was developed to help track students’ progress.
Usability Testing, Prototype
Coordinating the end-to-end usability testing process involved creating participant surveys, collaborating with engineers on the prototype, writing scripts, moderating tests, and preparing reports.
Usability Testing, Report
Usability Test Summary: We tested with an average of 5-6 kids, aged 5. The goals were to see if the kids understood how to play, and if they experienced boredom or frustration.
Week 5 Testing Results
Learnings: The game was easy to learn and engaging for the kids. However, the game progress indicator was challenging, and the adaptive engine felt narrow.
Recommendations: Rethink game instructions to be more age-appropriate. Improve adaptive rules to provide a more personalized experience.
Outcomes
The Adaptive Math Games project was the foundation for the My Math Academy, My Reading Academy, and My Reading Academy (Spanish) product lines, now serving over 50 million children worldwide.
As of 2025, Age of Learning’s annual revenue is estimated at $75 million, following a $300 million funding round in 2021 that brought the company’s valuation to $3 billion.
Lessons Learned
Balancing Intrinsic and Extrinsic Motivations: Intrinsic motivations, like the rewarding feeling of accomplishing tasks, combined with extrinsic motivations, such as earning points and rewards, were key to keeping kids engaged in the games.
Frequently engaging with real users matters: Ongoing conversations with real users provided invaluable insights to refine the games and meet their needs.
User testing is viable when operationalized. Efficient user testing can be implemented with a clear process and defined roles, allowing for fast feedback loops and iterations.