CAMS- and PMP-certified AML & KYC consultant with five years across KPMG, IBM Promontory, Deutsche Bank and BNY Mellon. I build working tools for fraud, transaction monitoring, sanctions and crypto investigations, so you can see how I think through a case.
Visit my portfolio → · LinkedIn · Email · Open to fintech, AML and risk roles (remote or New York)
| Project | What it shows | Links |
|---|---|---|
| Fraud Intelligence & Control Lab | Python + SQL feature engineering, control tuning with precision/recall tradeoffs, shared-device link analysis, executive recommendation | Live demo |
| Payments Strategy Lab | Payments unit economics, approval rate vs. fraud loss, review staffing and a 90-day pilot memo | Live demo |
| SQL Financial Crime Lab | Real SQLite in the browser: joins, CTEs, window functions, AML rule tuning and missed-case measurement | Live demo |
| Sanctions Review Desk | OFAC name search, identifier comparison, export-control context, information requests and QC | Live demo |
| Adverse Media Review Desk | Identity resolution that treats missing data as unknown, SQL triage views on SQLite / D1 | Live demo |
| Trace Desk AML Workspace | Live Ethereum USDT tracing, evidence pinning and portable case files | Live demo |
| Trace Desk: USDT Investigator | Reproducible 15-address Ethereum case: evidence, relationship graph and analyst report | Live demo |
- Investigations: transaction monitoring, SAR narratives and QC, KYC/CDD/EDD, beneficial ownership, PEP, sanctions and adverse media
- Data: SQL (SQLite, joins, CTEs, window functions), Python, Excel, control and rule tuning
- Platforms: NICE Actimize, LexisNexis, CLEAR, Thomson Reuters, DDIQ, Moody's RDC
- Delivery: PMP-certified; trained 20+ analysts, Agile project management, UAT, bilingual English/Spanish
A name match is a lead, not a conclusion. Missing information isn't a mismatch. Every decision should be explainable to the next reviewer. Each project documents its scope and limits, and uses fictional or public data.