cpdeol
← Back to Projects
Compliance + AI

Compliance & Risk Monitoring Platform

Achieved 98% money laundering detection accuracy and saved $50M+ in identified suspicious patterns — with full regulatory audit trail.

98%

detection accuracy

0

false positives on high-confidence alerts

$50M+

in suspicious patterns identified

75%

reduction in compliance review time

Role

Lead Business Systems Analyst — Compliance

Timeline

Q1–Q2 2024 · 5 months

Delivery context

AMLComplianceSQLFINTRACRisk Management

The Problem

Legacy AML systems flagged 40% of transactions as suspicious — an unsustainable false-positive burden on compliance teams. The organisation needed a system that compliance officers and regulators could trust, which required explainable alert reasoning alongside detection accuracy.

My Contribution

I led requirements analysis for this compliance monitoring platform, translating AML regulatory requirements into functional specifications for the detection and alerting system. I mapped the data flows from all source systems to the compliance dashboard, defined alert classification rules with compliance leadership, and designed the regulatory reporting data model to ensure audit trail completeness. I facilitated workshops with compliance officers to validate detection thresholds and alert routing logic, authored the UAT plan with real suspicious pattern scenarios, and coordinated integration with the regulator-facing reporting layer to ensure the system met audit documentation standards.

The Solution

Requirements-led compliance platform: AML regulation-to-specification translation, data flow mapping across source systems, alert classification definition, regulatory reporting data model design, and structured compliance team UAT before go-live.

Results

  • 98% detection accuracy on real money laundering cases
  • $50M+ in suspicious patterns identified
  • Zero false positives on high-confidence alerts
  • 75% reduction in compliance team review time
  • Full explainability for regulatory audits
Key learning
Compliance officers need to be able to explain every alert to a regulator. The explainability requirement wasn't a nice-to-have — it was the core acceptance criterion. Defining that in the requirements phase shaped the entire system architecture and meant the platform produced reports regulators could act on, not just statistics they had to interpret.

Tech Stack

Compliance

AMLFINTRACRegulatory Reporting

Data

SQLSSMSData ProfilingETL

Tools

JIRAConfluenceSharePoint

Methodology

Agile/ScrumBRD/TDDUATData Modeling

Related

How this project connects to the rest of my work.