from d -7 latency to real -time: the convergence of data mesh and event -driven architectures (EDA) in the orchestration of scalable financial ecosystems

Abstract

This article proposes a robust architectural model for large financial corporations facing the challenge of migrating from monolithic Data Warehouses and batch processes to federated ecosystems based on the Data Mesh paradigm. The central focus of the ana lysis lies in the drastic reduction of latency in data ingestion and consumption, addressing the transition from seven -day processing windows (D -7) to real -time (Online). Through an analytical approach grounded in software engineering and distribut ed systems theory, it investigates how the implementation of Event -Driven Architectures (EDA) acts as the necessary connective tissue to enable the decentralization proposed by Data Mesh. The case study presents the reverse engineering of a real digital tr ansformation, discussing the technical challenges of eventual consistency, computational governance in data meshes, and the strategic centralization of event consumption as a catalyst for the Customer 360º vision. The research concludes that the integration between autonomous data domains and continuous information f lows is the only sustainable path for competitiveness in the modern financial sector.