A collection of frameworks, case studies, articles, and operational work built across a career in fraud operations and compliance. More pieces added as they become available to share.
Projects
AI & Automation
Platform Integrity Investigation Agent — 48% reduction in manual effort
Designed and built an AI-powered investigation agent that performs forensic donation analysis and OSINT research with human-in-the-loop verification — cutting manual investigation time nearly in half within months of deployment. Includes a 7-stage OSINT pipeline, risk scoring framework, escalation logic, and full analyst sign-off workflow. Built as a team of one with no prior tooling in place.
Risk Scoring & Decision Frameworks
NPO Risk Scoring Calculator
An interactive risk-based scoring framework for evaluating nonprofit legitimacy and donation integrity in CSR programs. Assesses five categories — investigation request, platform activity, financial health, media profile, and public trust — and maps findings to a tier with a recommended action. Includes auto-escalation logic for the most critical cases.
ML & Data
ML Fraud Model — Data Validation & Tripwire Development
What happened when a model trained on a three-party CSR ecosystem flagged ordinary corporate giving as fraud, and the tripwires built once it was fixed: earmarked donation detection, volunteering rewards abuse, and IRS filing mismatches.
Articles & writing
Career & Mindset
Confessions of an Accidental Fraud Analyst
Nobody grows up wanting to be a fraud analyst. A candid, occasionally funny look at how a career built from support tickets, chargebacks, and NPO investigations turned into an actual mindset, and what it takes to build one yourself.
Fraud Operations
What NPO Fraud Actually Looks Like
A practitioner's guide to the tactics that appear most often in CSR fraud — from match harvesting and impossible volunteer hours to AI-generated documentation and trend-farming crisis NPOs.
AML & Strategy
FRAML: Why Fraud and AML Need to Work as One
Fraud and AML have historically operated as separate disciplines — but as financial crime evolves, so must the teams fighting it. A look at the case for convergence, the blindspots siloed programs create, and what a unified financial crime function actually looks like in practice.