Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
March 21, 2026Intelligent Systems with ApplicationsOpen Access

TRIAG: Tri-Reinforced Infused Generative Agents for Financial Risk Compliance

View Full Paper
Ask AI
Bookmark
Share

Authors

RSRafsun SheikhSMShah Jahan Miah

Discussion

Loading...

Member takes

Overview

Framework automates financial regulatory compliance tasks, enhancing efficiency for FinTech organizations, indicating improved decision-making.

Key Points

  • The aim is to develop and evaluate the TRIAG framework for better handling of financial regulatory compliance using generative AI agents.
  • Designed a novel framework using multi-agent reinforcement learning (MARL).
  • Integrated three LLM-based agents (Alpha, Beta, Gamma) for domain-specific expertise.
  • Utilized design science research paradigm for artifact creation.
  • Conducted qualitative and quantitative benchmarking with FinTech compliance experts.
  • Achieved a regulatory retrieval F1-score of 0.93.
  • Demonstrated a 96% reduction in inference costs.
  • Enhanced the efficiency and accuracy of compliance decision-making for officers.

Cite This Study

Sheikh et al. (2026) studied this question.

synapsesocial.com/papers/69be34d16e48c4981c672f3dhttps://doi.org/10.1016/j.iswa.2026.200653
View Full Paper
Ask AI
Bookmark
Share