Primary education
Word Explosion: an AI literacy and expression assistant
A learning PWA connects character recognition with words, sentences, stories and spaced review.

A learning PWA connects character recognition with words, sentences, stories and spaced review.
Correctly identifying a character does not by itself help a child use it in everyday expression. Children also need support when they cannot yet read or write it.
Speech, handwriting and stepwise learning support an adaptive queue. Story versions, parent reports, multiple child profiles and reviewed teaching content keep learning and family oversight connected.
Learning behavior informs guided feedback, connecting teaching content, learning queues and parental observations while implementing access controls and data boundaries for each child’s profile.
AI methods
CHINESE_LITERACY_LEARNING / ADAPTIVE_REVIEW_QUEUE / STORY_EXPRESSION_COACH / PARENT_GROWTH_REPORT / EDUCATION_CONTENT_GOVERNANCE
Workflows, pipelines and agents
Children's learning PWA / Curriculum vocabulary and reviewed teaching content / Initial assessment, adaptive learning queue and spaced review / AI expression feedback and story versions / portfolios / Parent reports, multiple child profiles and access isolation / Supabase / PostgreSQL / RLS data services
Curriculum and vocabulary require authorization, review and manual proofreading. Family accounts, child profiles, learning records and works are isolated by account and role. AI keys remain on the server. Voice, text and generated content use layered retention policies informed by privacy and safety assessment.
Moments in a child's learning journey
Start with one character and follow learning over time


Children can use new characters in their own stories, while parents can follow the development of their expression.
This public project account describes documented work and its current stage. It does not imply independent verification, client endorsement or unreported commercial results.