AI-ENHANCED FRAUD DETECTION IN FINANCIAL WORKFLOWS: A HYBRID ML-LLM FRAMEWORK FOR RISK SCORING AND ANOMALY ANALYTICS

Artem Tsymbal

doi:10.30890/2567-5273.2025-42-03-004

AI-ENHANCED FRAUD DETECTION IN FINANCIAL WORKFLOWS: A HYBRID ML-LLM FRAMEWORK FOR RISK SCORING AND ANOMALY ANALYTICS

Authors

Artem Tsymbal National Technical University of Ukraine “Igor Sikorsky Kyiv Polytechnic Institute” https://orcid.org0009-0006-8786-8428

DOI:

https://doi.org/10.30890/2567-5273.2025-42-03-004

Keywords:

fraud detection, risk scoring, financial workflows, machine learning (ML), large language models (LLM), real-time analytics, stream processing, semantic enrichment, explainable AI (XAI), latency and SLO, compliance, auditability.

Abstract

The rapid expansion of cashless payments and instant transfers has intensified performance and transparency requirements for fraud detection. Decisions to block or approve transactions must be issued within milliseconds, without compromising accuracy, use

Downloads

Published

2025-12-30

How to Cite

Цимбал, А. (2025). AI-ENHANCED FRAUD DETECTION IN FINANCIAL WORKFLOWS: A HYBRID ML-LLM FRAMEWORK FOR RISK SCORING AND ANOMALY ANALYTICS. Modern Engineering and Innovative Technologies, 3(42-03), 3–17. https://doi.org/10.30890/2567-5273.2025-42-03-004