Event-Driven Microservices for Scalable Omni-Channel Retail: A Serverless Architecture Framework

Abstract

Omni-channel retail relies on systems which need to coordinate in real-time between the digital store, the physical store, the warehouse, the payment system, the customer engagement platforms and the last-mile fulfilment networks. The traditional monolithic and tightly coupled service design doesn't align well with rapidly changing demand, heterogeneous channel interactions, real-time visibility of inventory, and failure isolation requirements. The combined use of event-driven microservices and serverless computing provides a promising architectural basis for elastic, loosely coupled and operationally resilient retail platforms, but there is still a technical fragmentation in the literature of software architectures, distributed systems, retail operations and retail solutions in how the two are applied. This review looks at peer reviewed evidence of microservice decomposition, publish subscribe communication, complex event processing, serverless execution, distributed consistency and omni channel fulfilment. The review proposes an architectural framework that uses canonical business events, serverless execution for burst workloads, event brokers for asynchronous workflows, and read-optimized projections for channel-specific experiences. The review suggests that granularity, inventory consistency, idempotency, observability, and failure compensation are important design principles when treated as first-class architectural concerns. Reproducible benchmarks and reference datasets remain limited, and further work is needed on cold-start mitigation and event-schema governance for channel-specific retail workloads. Further research should be conducted with empirical evaluation of serverless event pipelines with expected event traffic from all channels and formal consistency models for inventory events and standardized quality measures for serverless architecture.