ComplianceNLP system uses RAG and knowledge graphs to detect regulatory gaps

Researchers have developed ComplianceNLP, a system designed to automate the monitoring of regulatory changes and identify compliance gaps for financial institutions. The system utilizes a knowledge-graph-augmented RAG pipeline, processing over 12,000 regulatory provisions from frameworks like SEC, MiFID II, and Basel III. In testing, ComplianceNLP achieved an 87.7 F1 score for gap detection, outperforming GPT-4o+RAG, and demonstrated significant efficiency gains in a real-world deployment. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT Automates regulatory monitoring and gap detection, potentially saving financial institutions significant time and reducing fines.

RANK_REASON Academic paper detailing a new system for regulatory compliance.

Read on arXiv cs.CL →

• ComplianceNLP
• SEC
• MiFID II
• Basel III
• LEGAL-BERT
• Medusa
• GPT-4o

• paper
• product
• infra