Red Store - Enterprise Point of Sale & Inventory Management System

A distributed, real-time Point of Sale and inventory management system built for retail environments, featuring multi-device synchronization, event-driven architecture, and high-performance caching. These demo videos were recorded by my friend to educate the store staff on how to effectively use the barcode scanner with the overall software

Demo 1

Watch Demo 1

Demo 2 (OTP system was in production already, so we force fed the OTP code through console for only development)

Watch Demo 2

Demo 3

Watch Demo 3

System Architecture

Red Store is built as a microservices architecture with event-driven communication, designed to handle concurrent operations across multiple iPad terminals while maintaining data consistency and real-time synchronization.

Core Services

• Streaming Service - Apache Kafka event streaming platform with custom partitioning
• Consumer Service - Message processing and database operations
• Producer Service - Sales and inventory event publishing
• Cache Layer - Redis-based caching and pub/sub messaging
• Database Layer - PostgreSQL with TimescaleDB for analytics

Key Features

Real-Time Multi-Device Synchronization

• Concurrent sales processing across multiple iPad terminals
• Instant inventory updates with zero conflicts
• Server-Sent Events (SSE) for real-time UI updates

Event-Driven Architecture

• Apache Kafka message streaming with custom partitioning logic
• Automatic topic discovery and consumer group management
• BullMQ job queues for asynchronous processing

High-Performance Caching

• Redis pipeline operations for optimized performance
• Pub/Sub messaging for instant synchronization
• Strategic cache invalidation and updates

Analytics & Reporting

• TimescaleDB integration for time-series data
• Real-time sales analytics and reporting
• Historical transaction tracking

Technology Stack

Backend

• Runtime: Node.js with TypeScript
• Framework: Express.js with REST APIs
• Message Queue: Apache Kafka
• Cache: Redis with pub/sub
• Job Queue: BullMQ
• Database: PostgreSQL (Supabase) + TimescaleDB

Architecture Patterns

• Microservices Architecture
• Event-Driven Design
• Publisher-Subscriber Pattern

Multi-Device Setup

The system is designed for iPad terminals but supports any web-capable device:

1. Configure store-specific partitioning
2. Set up device-specific consumer groups
3. Enable real-time SSE connections
4. Configure Redis pub/sub channels

Security Features

• Store-specific data isolation through partitioning
• User-based topic segregation
• Redis channel-based access control
• Database row-level security policies

Scalability

The system architecture supports:

• Horizontal scaling through Kafka partitioning
• Multiple consumer groups for load distribution
• Database sharding through store-based partitioning

These pictures show the on-ground implementation of the Red Store system at the store.
They capture how the system was setup to be used by the staff.

---

Store Setup – Picture 1

---

Store Setup – Picture 2