Archive

The 12 Ways AI Agents Fail That Your Logs Will Never Show You

How Risk Committees Interpret AI Outputs

SS21 - Translating technical dashboards and model exceptions into governance-friendly narratives

Unsupervised Learning for AML Pattern Detection

SS20 - How clustering logic supports investigations without labeled fraud data

Security Architecture for AI Systems

SS19 - Threat Models, Isolation, and Why Your Model Is More Dangerous Than Your Database

Fairness & Demographic Slice Monitoring

WS18 - How Banks Actually Measure (and Fix) Discriminatory AI Without Drowning in Metrics

Deployment Gate Pack Generator

WS17 - The Documentation Checklist That Gets Models From "Ready" to Approved in Production

Fairness Metrics and Bias Ratios in Finance

SS18 - Why 95% Accuracy Can Be Discriminatory and How to Measure Fair Lending

How Risk Committees Interpret AI Outputs

SS17 - Translating Model Performance Into Decisions That Get Approved

AI Incident Response Runbook + Escalation Matrix

WS16 - When Your Model Breaks in Production, Who Calls Who and What Do They Do?