Fraud Detection & Financial Crime Pattern Intelligence

Harwick Building Society · £4.2B monthly transaction volume

Net Fraud Loss Rate

11.2 bps

+93% vs 18mo

Alert Precision Rate

18.4%

-34% vs Q1

Detection Lag

6.8h

+2.1h vs target

SAR Filing Timeliness

100%

On target

Case Closure in SLA

76%

-9pp vs Q1

Network Detection Share

34%

+12pp vs Q1

Typology Velocity Heat Map — 12-Week Trend

Alert Volume vs. Precision Rate

Top 5 False Positive Generators

Active Case Intelligence — Priority Queue

Case ID	Typology	Amount	Age (d)	Analyst	Network Flag	Status
FC-28491	APP Scam	£47,200	2	L. Chen	Yes	Investigation
FC-28483	Card Not Present	£3,890	4	M. Patel	No	Triage
FC-28476	APP Scam	£21,500	6	R. Johnson	Yes	Escalated
FC-28469	ACH Takeover	£12,800	8	S. Kumar	Yes	Investigation
FC-28457	Wire Fraud	£89,400	11	L. Chen	No	Investigation
FC-28441	APP Scam	£34,100	13	M. Patel	Yes	SAR Filed
FC-28428	Check Fraud	£6,750	15	R. Johnson	No	Triage
FC-28412	Smurfing	£18,900	18	S. Kumar	Yes	Investigation

Rule Decay Index — Age vs. Precision

Recovery Rate by Typology

Investigator Utilization

87% capacity · 2.1 cases/analyst

Average Case Resolution Time

11.3 days · Target: 8 days

Queue Backlog Growth

+18% vs 30d ago · 142 open cases

Net Fraud Loss Rate Trend with Q3/Q4 Projection

Projection scenario: With improved alert precision (target 25%) and detection lag reduction (target <4h), model forecasts 7.2 bps by Q4 end — within goal threshold. APP scam mitigation is the critical path: reducing APP share from 41% to 28% of mix would deliver 2.8 bps improvement alone.