Graduation project
IntelliPharma
A bilingual, AI-driven ERP and CRM for pharmaceutical distribution.
IntelliPharma is my graduation project — a bilingual pharmaceutical ERP and CRM serving warehouse admins, sales reps, and distributors, built by a team of five around a reinforcement-learning route planner as its headline feature. I owned the frontend end to end: a strict Feature-Sliced Design codebase spanning 1,100+ source files, 21 domain entities, and 67 route-level pages.
The frontend rests on a two-layer data architecture — a typed Axios client with normalized error handling, and reusable TanStack Query hooks for Suspense reads, infinite lists, and mutations. Backend-authoritative role-based access control protects routes, navigation, and UI down to individual capabilities, and a companion Flutter app and Laravel backend round out the system.
Highlights
- Architected a strict Feature-Sliced Design codebase across app, pages, features, entities, shared, and widgets layers, keeping dependency direction explicit across 1,100+ source files.
- Designed a two-layer data architecture: a typed Axios apiClient with interceptor-normalized responses, plus reusable TanStack Query hooks for Suspense reads, infinite lists, and CRUD mutations.
- Implemented backend-authoritative RBAC from session state through protected routes, permission-aware navigation, and capability-driven UI that fails closed when access is missing.
- Integrated real-time field-team tracking over authenticated WebSocket channels, reconciling REST snapshots with live location events in a normalized store.
- Engineered cross-tab authentication as a concurrency system — serialized token refreshes with Web Locks and propagated session state across tabs via BroadcastChannel.