Restaurant Management & POS Platform

Full-Stack Developer, Caisse Manager

Caisse Manager builds a SaaS platform for restaurant management and point of sale, used in live restaurant operations. The system spans a React front end, a Node.js/Express backend, MongoDB, Redis, and a containerized deployment behind Nginx.

REST APIs serve the point-of-sale and management front end; Socket.IO carries real-time updates; Redis backs both caching and the Bull-based job queue; the whole stack runs as multiple Docker services coordinated behind Nginx.

As a production system used in real restaurant operations, the platform needed reliable APIs and real-time updates, background operations that don't silently fail, and a way to trace performance or security problems back to their root cause without destabilizing the rest of the system.

The backend is one part of a multi-service, containerized architecture — a change in one service can affect others that depend on it. That made impact analysis, not just the fix itself, a core part of the job.

• Developed and maintained full-stack features for the platform using React, Node.js, Express.js and MongoDB.
• Designed and maintained REST APIs, business integrations and real-time mechanisms using Socket.IO and Redis.
• Built asynchronous processing and task queues with Bull for reliable background operations.
• Analyzed service performance and stability to identify the technical causes of slowdowns, errors and abnormal behavior.
• Investigated and fixed security issues related to authentication, tokens, API access and data migrations.

• Used Redis for both real-time state and as the backing store for Bull, keeping the caching and queueing layers on shared, well-understood infrastructure.
• Ran the stack as separate Docker services behind Nginx so components could be reasoned about and deployed independently.
• Treated authentication, token handling and API access as an ongoing security surface to review, not a one-time setup.

Node.js / Express

REST API layer serving the POS and management front end.

React

The point-of-sale and restaurant-management front end.

MongoDB

Primary data store for orders, products and business records.

Redis

Real-time state and the backing store for the Bull job queue.

Socket.IO

Pushes real-time updates to connected clients.

Bull

Background/async job processing that doesn't block the request path.

Docker

Each service runs as its own container, deployed independently.

Nginx

Routes and fronts the containerized services.

Contributed to a more stable, better-documented production system — with clearer diagnostic procedures shared across the technical and operational teams, and known authentication/token gaps closed.

Working inside a live, multi-service production system puts a premium on tracing impact before making a change — the fix is often the easy part.