Explainable Anomaly Detection in Multimodal Telemetry of Banking Microservices

Natalia Kalivoshko

Citation: Natalia Kalivoshko, "Explainable Anomaly Detection in Multimodal Telemetry of Banking Microservices", Universal Library of Business and Economics, Volume 02, Issue 04.

Copyright: This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

The rapid proliferation of microservice architectures in the banking sector has introduced complex operational challenges related to real-time fault detection across heterogeneous telemetry streams, including metrics, logs, distributed traces, and business event signals. Existing anomaly detection systems in financial environments often lack interpretability, limiting operator trust and regulatory auditability. This study investigates the application of explainable artificial intelligence (XAI) methods, specifically SHapley Additive exPlanations (SHAP), Local Interpretable Model-agnostic Explanations (LIME), and attention-based neural architectures, integrated with unsupervised multimodal anomaly detection models for banking microservice telemetry. A systematic comparative analysis of detection approaches is conducted, supported by case study evidence from a Central Bank Digital Currency (CBDC) integration pilot in a systemically important Russian bank. The results demonstrate that the proposed framework achieves a detection accuracy of 0.91, a false positive rate of 0.09, and a mean time-to-detect (MTTD) reduction of 74 % on average compared to legacy black-box methods. The study concludes that combining multimodal correlation engines with operator-readable XAI outputs constitutes a viable and necessary evolution of AIOps practices in regulated financial institutions. The findings are relevant for architects, reliability engineers, and compliance officers in the financial services industry.


Keywords: Explainable Artificial Intelligence, Anomaly Detection, Multimodal Telemetry, Banking Microservices, SHAP, LIME, Aiops, Distributed Tracing, Observability, CBDC Integration.

Download doi https://doi.org/10.70315/uloap.ulbec.2025.0204015