An AI reading platform for UK primary school children. Personalised stories generated at the right reading level, VIPERS comprehension questions, progress tracking, SATs prep, and dashboards for parents and teachers. Built solo, end to end.
Parent dashboard — 4 children, 3 reading levels, 139 topics, 100% safety rating
Children reading at home face two problems: finding books at the right level is hard, and comprehension practice feels like homework. Primary Story needed to make reading genuinely engaging while still building the skills tested at KS1 and KS2.
Primary Story generates personalised stories on demand — the child picks a topic, length, and character, and the AI produces a story calibrated to their reading level. Every story comes with comprehension questions built around the VIPERS framework used in UK primary schools (Vocabulary, Inference, Prediction, Explain, Retrieve, Summarise).
Building for children required a level of care that most AI products don't need. Content safety guardrails, age-appropriate language calibration, and a UX that a seven-year-old could navigate independently — all while satisfying the parents and teachers who actually control the subscription.
The multi-role auth model (child account, parent account, teacher account with class management) required careful data separation and permission logic at the database level, not just the UI.
Azure OpenAI was chosen over the standard OpenAI API to meet data residency requirements for a product handling children's data under UK GDPR.
I've built child-safe AI content generation, multi-role dashboards, and curriculum-aligned tools for production use. Available as a fixed-price sprint — one feature, 7 working days, from £750.