Project Description
La Jefecita is a full-stack restaurant management platform built to replace paper tickets and walkie-talkie communication between waitstaff, kitchen, and cashiers. The system unifies four role-specific interfaces (waiter, cook, cashier, admin) behind a real-time WebSocket layer that keeps order state consistent across every device on the floor.
Business Impact
- ~30% reduction in order-entry errors by replacing handwritten tickets with role-scoped digital flows and instant kitchen confirmation.
- Ticket turnaround cut from minutes to seconds: WebSocket-driven order updates eliminated the kitchen ⇄ cashier relay delay and removed the most common cause of late dishes.
- Eliminated end-of-shift cash discrepancies by closing the loop between order, payment (Stripe), and cashier reports — every paid ticket is reconciled automatically.
- Single source of truth for menu, pricing, and stock across roles, removing the cost of out-of-sync printed menus during promotions.
Technologies Used
- Frontend: Next.js (TypeScript)
- Backend: NestJS (TypeScript)
- Real-time layer: WebSockets (Socket.IO)
- Payments: Stripe (sandbox simulation)
- Database: PostgreSQL
- Security: JWT-based role authentication & per-role authorization guards
Key Features
- Role-Specific Management: Dedicated UIs for waitstaff, cooks, cashiers, and admins, each scoped to the exact actions and data that role needs.
- Real-Time Order Sync: WebSocket events push order state changes (placed, in prep, ready, served, paid) to every relevant screen in <300 ms.
- Secure Payment Integration: Stripe-backed checkout flow with auditable transaction logs.
- JWT Auth & Authorization: Guards enforce role boundaries at the API layer, not just the UI.
Conclusions
La Jefecita proved that even a small restaurant operation gets outsized ROI from a focused real-time stack: fewer errors per shift, faster table turns, and clean financial reconciliation. The architecture (NestJS API + Next.js clients + WebSocket gateway + Postgres) scales cleanly to multi-location deployments.