EXPLORE PUBLICATIONS BY COUNTRIES

Event-Driven Architecture, Retail Transactions, Cloud Messaging, Reactive Frameworks, Microservices.

Modern retail platforms must support high-throughput, low-latency transaction processing under unpredictable and bursty workloads. Traditional monolithic architectures frequently fail to meet these demands due to blocking interactions, limited horizontal scalability, and susceptibility to cascading failures. This study proposes a cloud-native Event-Driven Architecture (EDA) that integrates Amazon Simple Queue Service (AWS SQS) with reactive Spring WebFlux microservices to enable asynchronous, fault-tolerant, and elastically scalable retail transaction processing. The following research question guides this work: RQ—Can an event-driven, message-queue–based architecture combined with reactive microservices significantly improve scalability, latency, and fault tolerance in real-time retail transaction pipelines compared to a monolithic baseline?

To answer this question, I develop a full prototype of implementing order creation, inventory reservation, payment authorization, and customer notification. Using a heterogeneous dataset of 412 simulated retail transactions, I conduct extensive load tests at 100, 500, and 1000 concurrent users. Results show that the EDA system achieves near-linear throughput scaling, maintains low and predictable latency, and isolates service failures without degrading overall system performances significantly outperforming the monolithic benchmark. These findings demonstrate that EDA, supported by cloud-managed queues and reactive runtimes, provides a viable and robust architectural foundation for next-generation retail platforms that require elasticity, resilience, and real-time responsiveness.