Nori

Nori: An Empathy-Driven Adaptive Learning Journey

Inspiration & Research

Nori was inspired by the emotional struggles of families dealing with neurodivergent challenges such as dyslexia and ADHD. Traditional achievement-based games often rely on parental pressure and rigid metrics, which can heighten learning anxiety. Our research, grounded in Social Emotional Learning (SEL) and Scaffolding Theory (Bruner, Vygotsky), found that performance-based systems often reduce intrinsic persistence. Nori was conceived to replace these extrinsic pressures with an altruism-driven model where knowledge becomes a tool for compassion.



Project Description & Unique Properties

Nori is an empathy-driven adaptive educational game that cultivates intrinsic motivation by reframing learning as an act of care. Instead of chasing scores, children help and protect animals within a narrative world. Globally, this addresses an urgent need for the 20% of children with neurodivergent traits, including approximately 160–380 million with dyslexia and 85–130 million with ADHD. By integrating academic subjects—math, spelling, and science—into a unified emotional framework, Nori supports diverse learning needs through a barrier-free, inclusive UX.



Operation & Interaction

Children explore diverse biomes, such as islands or volcanoes, solving challenges to rescue animals. In return, grateful animals share educational "fun facts" and keepsakes for the child’s digital room. A dedicated Collection page allows children to revisit their impact. To ensure a healthy digital-physical balance, when screen time ends, an AI-powered IoT plush companion takes over, offering personalized stories and quizzes based on the child’s favorite in-game moments.



Technical Realization

Developed in Unity 6 with a Real-time Cel-shading pipeline for a friendly aesthetic, Nori leverages the Gemini LLM API and OpenCV-based Affective Computing for empathetic, non-linear interactions. Supported by Google Firebase, the system tracks "kindness achievements" in real-time. Designed for iPadOS (2732 x 2048 px), the project utilizes AI scaffolding to adjust difficulty dynamically, ensuring that learning remains a meaningful pathway toward helping the world.