Project Overview
A UK FCA-regulated payments company processing 2.4M transactions per month needed to replace manual transactio...
Technology Stack
Compliance & Standards
The Challenge
A UK FCA-regulated payments company processing 2.4M transactions per month needed to replace manual transaction monitoring (a team of 3 compliance analysts reviewing flagged transactions 8 hours per day) with an AI-powered system. Requirements: FCA JMLSG (Joint Money Laundering Steering Group) alignment, real-time transaction scoring, PEP/sanctions screening, SAR (Suspicious Activity Report) generation support, and FCA-auditable decision trail. Budget: £160,000.
Our Approach
ML Transaction Scoring Engine
Python scikit-learn gradient boosting model trained on 18 months of historical transaction data (1.2M labelled transactions — compliant, suspicious, confirmed fraudulent).
Features
- transaction amount, merchant category, geographic pattern, velocity (transactions per hour), peer comparison (deviation from cohort behaviour), device fingerprint, and time-of-day pattern.
- Model deployed as FastAPI microservice.
FCA JMLSG Alignment
Three-tier risk categorisation: Low (automated pass), Medium (enhanced monitoring — automated + analyst review queue), High (SAR candidate — mandatory analyst review before any automated action).
JMLSG guidance
- no fully automated adverse decisions on customers without human review.
- All High-risk flags routed to human analyst within 4 hours.
OFSI/HMT Sanctions and PEP Screening
- Real-time sanctions screening: OFSI UK (Office of Financial Sanctions Implementation) consolidated sanctions list, HM Treasury UK sanctions lists, and Dow Jones World-Check for PEP (Politically Exposed Persons) screening.
- Fuzzy matching algorithm (Levenshtein distance + phonetic matching) to catch name variations and transliterations.
Every transaction score logged
- model version, feature values at time of scoring, score, risk tier, human review outcome (if any), and SAR reference (if generated).
- Immutable audit log (AWS CloudTrail + S3 Object Lock) — the FCA can reconstruct any decision at any point in time.
The Results
System live at 14 weeks, £148,000 — under budget.
False positive rate: 68% reduction (compliance analyst time focused on genuine risk).
Transaction scoring latency: 47ms average (real-time at transaction).
SAR quality: FCA acknowledged improvement in SAR specificity in annual visit.
Compliance analyst capacity: freed from 8 hours/day review to 2 hours/day (remaining cases are genuinely complex).
FCA Consumer Duty: automated decisions logged and human override available per Article 22.
“The 68% false positive reduction was the number that convinced our board. Our compliance team was spending 80% of their time on noise — the AI cut that dramatically without reducing genuine risk detection. The FCA audit trail is exactly what our supervisors wanted to see." — Chief Compliance Officer, UK Payments Company (name withheld)”
Project Details
Related Case Studies
View AllOpen Banking Cash Management Platform — UK SME Lender
Open Banking AISP platform for UK SME lender. FCA COBS, PSD2, UK GDPR. £42M lending decisions improved. Case study.
FinTech Mortgage Decision Engine — UK Challenger Lender
A UK FCA-authorised mortgage lender offering residential mortgages needed to rebuild their mortgage decision e...
FinTech Trade Finance Platform — UK SME Export Finance
A UK alternative finance company providing trade finance (invoice discounting, supply chain finance, and lette...
FinTech Crypto Asset Platform — UK FCA Registered Business
A UK company registered with the FCA as a cryptoasset business (under MLRs 2017 Schedule 7A) wanted to build a...
Ready to Transform Your Business?
Let's discuss how our technical expertise can help you achieve remarkable results.