EvoAML: Evolutionary Anti-Money Laundering RegTech Framework

English | 中文

Overview

EvoAML is an open-source RegTech framework designed to bridge the gap between advanced AI research and practical anti-money laundering compliance. It provides a modular pipeline for detecting complex money laundering patterns across cross-industry financial networks.

Core Architecture

EvoAML integrates multiple detection methodologies to provide comprehensive coverage:

1. Graph-Driven Cross-Industry Tracking [In Progress] Uses graph-based analysis to trace financial flows across supply chains and multi-sector networks.

2. Temporal Evolution Analysis [In Progress] Applies time-series analysis to identify evolving behavioral patterns indicative of laundering.

3. AMLA 2020 Compliance Engine [Implemented] Translates detection results into standardized BSA-compliant SAR (Suspicious Activity Report) narratives.

Quick Start

1. Install dependencies: pip install -r requirements.txt
2. Run the dashboard: streamlit run app.py
3. See examples/ for algorithmic workflows.

Roadmap

• Phase 1: Foundation & Compliance Architecture ✅
• Phase 2: Data Ingestion & Preprocessing ✅
• Phase 3: Graph Tracking Module ✅
• Phase 4: Temporal Analysis Module ✅
• Phase 5: Compliance Engine ✅
• Phase 6: Visualization & Release ✅

Data Format

EvoAML accepts transaction data in CSV format. The following columns are required:

• sender (string): Source entity identifier
• receiver (string): Destination entity identifier
• amount (float): Transaction amount in USD

Optional columns:

• sector (string): Industry sector (e.g., Energy, Finance, Logistics)
• timestamp (string): Transaction date (ISO format preferred)

Example CSV

sender,receiver,amount,sector,timestamp
Entity_001,Entity_002,50000,Energy,2024-01-15
Entity_003,Entity_001,25000,Finance,2024-01-16
Entity_004,Entity_005,75000,Logistics,2024-01-17

You can upload this CSV directly in the Streamlit dashboard to run the analysis on your own data.