Latency-Aware Data Routing for Adaptive Query Optimization in MultiTenant Financial Systems

Abstract

Multi-tenant financial systems, including SaaS banks, neo-banks, and mortgage SaaS platforms, face significant challenges in sustaining low-latency query performance while accommodating heterogeneous workloads across geographically distributed infrastructures. This paper proposes a latency-aware data routing framework that integrates adaptive query optimization to dynamically select execution paths based on real-time network conditions, tenant-specific priorities, and regulatory compliance constraints. The architecture incorporates predictive latency modeling, topology-aware routing, and workload classification to consistently achieve sub-second response times in high-throughput financial environments. Experimental evaluations on both synthetic and real-world transaction datasets indicate latency reductions of up to 37% compared with baseline routing strategies, without compromising compliance, auditability, or service-level adherence.