Family-focused AI workspace with parent and child profiles, a local-first study system with FSRS review scheduling and background sync, and per-account usage and cost monitoring.
Role
Planning, design and full-stack development
Period
2025 – 2026
Status
Private and unreleased; my involvement ended in April 2026
How it fits together
This project is private, so there are no screenshots. Here's how the learning system is put together.
On the device
IndexedDB
Zustand store
FSRS scheduler
Immutable review events
Background sync
with stale-write protection
Supabase / PostgreSQL
Profiles & permissions
Review history
Usage & cost monitoring
What I did
Designed and built separate parent and child profiles, chat-style AI assistant flows, guided content editing, usage tracking and admin oversight.
Designed a local-first learning subsystem combining IndexedDB, Zustand and Supabase / PostgreSQL.
Implemented immutable review events, background synchronization and protection against stale writes.
Implemented review scheduling with the FSRS algorithm (ts-fsrs).
Built account-level monitoring of conversations, model usage, token consumption and estimated cost.