Senior Data Science & Analytics Leader | Principal Architect for $1B+ Remediations & Federal Consent Order Execution | Forensic Data Engineering | Strategic Management Advisor | Board-Ready Insights | “No Cold Handoffs”
Principal Data Strategist and Senior Data Science Leader with 16 years of experience architecting enterprise-scale decision systems for mission-critical challenges. While my background includes navigating $1B+ mandates and federal regulatory closures, my core expertise lies in designing end-to-end systems that ensure data integrity, stress-test decision strategies, quantify financial impact, and drive growth outcomes. I architect audit-ready pipelines that translate complex data into actionable, board-ready insights for the C-Suite and stakeholders at every level.
-
credit_decisioning_strategy
Enterprise Credit Decisioning Strategy Simulator — Governed PostgreSQL credit decisioning simulator for synthetic applications, pre-production strategy testing, matched comparison, counteroffer governance, and Expected Loss tradeoffs without PII. -
forensic-data-integrity
Universal Data Integrity: Forensic Gatekeeper Engine — Python diagnostic engine for scoring raw dataset health, surfacing hidden nulls, dominant bias, format defects, and zero-variance fields, and converting data-quality risk into audit ledgers and executive scorecards. -
enterprise-reconciliation-reporting
Enterprise Reconciliation Reporting: Automated Audit Governance — SAS reconciliation framework for tolerance-aware A/B dataset comparison, schema drift detection, key-only and value-difference testing, run metadata, UAT validation, and CSV-ready audit outputs. -
metro2-remediation-sandbox
Metro 2 Remediation Sandbox: Synthetic Portfolio and Impact Engine — Modular PostgreSQL sandbox for synthetic longitudinal tradeline generation, credit-impact window evaluation, cure logic, treatment assignment, QA, and audit-ready before-and-after reporting. -
insurance-coverage-reconciliation
Insurance Coverage Reconciliation: Audit and Liability Cost Signal — Metadata-driven SAS engine for merging overlapping proof-of-coverage periods, comparing proof evidence against bank-issued policy windows, applying lapse-threshold rules, and calculating policy adjustment factors. -
financial-tvm-optimization
Financial TVM Optimization: Remediation Liability Engine — High-scale SAS TVM engine using 1-year CMT Treasury rates, PROC FCMP arrays, chunked processing, annual compounding, and final remainder interest for remediation liability calculations. -
survival-retention-engine
Survival Retention Engine: Growth Strategy and Runway Modeling — Python survival-analysis engine combining K-Means personas, CoxPH runway modeling, high-risk scenario simulation, Kaplan-Meier curves, and executive PPTX/PDF reporting.
My GitHub portfolio is organized as a public proof layer for enterprise analytics architecture. Each project demonstrates a different point in the lifecycle from raw-data trust to decision strategy, remediation, liability modeling, and growth optimization.
Data Integrity
→ Reconciliation and Signal Detection
→ Credit Decisioning Strategy
→ Remediation Simulation
→ Liability and TVM Quantification
→ Growth / Retention Strategy
The common design pattern across the portfolio:
Governed Inputs
→ Diagnostic Logic
→ Scenario or Strategy Layer
→ Validation Evidence
→ Executive Narrative
→ Audit-Ready Handoff
This is the same philosophy I apply in regulated enterprise environments: do not simply produce outputs; make the underlying logic, evidence, assumptions, and stakeholder interpretation clear enough to trust.
Enterprise Credit Decisioning Strategy Simulator
A two-module, portfolio-grade PostgreSQL credit decisioning simulator designed to demonstrate how governed synthetic application generation, pre-production strategy testing, archive-backed comparison, and executive-ready evidence can work together without exposing PII.
This is the newest addition to the portfolio architecture. It expands the lifecycle from forensic data integrity and remediation simulation into full credit strategy design, scenario testing, counteroffer governance, matched comparison, and executive decisioning evidence.
Builds a deterministic synthetic application population and risk foundation across:
- product type
- score band
- borrower profile
- requested exposure
- APR proxy
- monthly payment
- payment-to-income ratio
- synthetic estimated-PD proxy
- LGD
- Expected Loss
Campaign scope:
- 19 governed scenarios
- 50,000 applications each
- 950,000 archived scenario rows
- baseline product × score risk surface
- scenario lever matrix
- scenario Expected-Loss impact decomposition
Consumes Module 1 synthetic outputs and applies governed strategy controls to simulate:
- approvals
- counteroffers
- manual reviews
- declines
- approved exposure
- approved Expected Loss
- ordinary vs aggressive counteroffer path evidence
- matched baseline / challenger strategy comparison
- strategy frontier tradeoffs
Campaign scope:
- 39 governed strategy runs
- 50,000 applicants each
- 1.95M archived strategy decisions
- 20 matched comparison groups
- 7 scenario families
- strategy frontier analysis
- baseline-relative challenger tradeoffs
- selective vs aggressive counteroffer governance
This project demonstrates:
- governed credit strategy architecture
- synthetic application and risk modeling
- pre-production decision simulation
- matched population comparison
- archive-backed evidence discipline
- parameterized policy and product controls
- validation-ready QA workflows
- executive dashboard storytelling
- Expected Loss and affordability tradeoff analysis
- counteroffer feasibility and governance design
This is not a dashboard-only artifact. It is a full-system decisioning simulator with SQL implementation, documentation, synthetic samples, QA outputs, campaign evidence, and executive dashboards.
- Principal Architect for the end-to-end CPI Proof of Insurance intake.
- Engineered modular, memory-efficient SAS arrays to reconcile high-dimensional, disparate insurance data against historical burdens and provide TVM-adjusted customer redress.
- Led CPI Credit Bureau remediation, ensuring FCRA compliance and traceability.
- Translated Consent Order requirements into “No Cold Handoff” diagnostic dashboards and zero-defect regulatory reporting.
- Engineered centralized credit furnishing protocols adopted across the Wells Fargo enterprise.
- Architected automated surveillance for 100M+ transmissions to identify systemic patterns, root causes, and FCRA compliance risk.
- Established a global governance framework to protect immutable favorable reporting and refine targeted remediation furnishing, saving $60k per submission.
- Directed strategy, risk portfolio, and Lean Six Sigma DMADV lifecycle for 15+ Consumer Auto Finance mandates, including high-risk SCRA and CPI efforts.
- Served as senior management advisor for data fidelity, risk response, operational execution, and regulatory-grade decision support.
- Modernized legacy analytical pipelines by establishing modular code libraries and best-practice SQL/SAS/Python frameworks.
- Engineered Cox Proportional Hazards and Logistic Regression models to quantify longitudinal default likelihood for 5+ published research papers.
- Defined account archetypes and analyzed covariate hazard ratios to isolate the impact of macro-economic triggers on customer runway.
- Simulated baseline hazard shifts and provided OCC leadership and the National Risk Committee with multi-layered Tableau dashboards detailing projected risk-reduction metrics.
- Bachelor of Arts in Economics, Minor in Mathematics — Virginia Tech University
- Lean Six Sigma Yellow Belt — Office of the Comptroller of the Currency
- Associate Citation in Project Management — George Washington University
- Data Architecture: SQL, PostgreSQL, SAS, Python, Snowflake, Microsoft Fabric, SQL Server, Teradata
- SAS Engineering: SAS Macro, PROC FCMP, SAS Dictionary Tables, 2D Arrays, Metadata-Driven Processing, Regulatory-Grade Reporting
- Python Analytics: Pandas, NumPy, Scikit-learn, Lifelines, Matplotlib, OpenPyXL, python-pptx, ReportLab
- Decision Systems: Credit Strategy Simulation, Policy Rule Engines, Scenario Testing, Matched Comparison, Strategy Frontier Analysis
- Data Science: Survival Analysis, Cox Proportional Hazards, Logistic Regression, K-Means Clustering, Pattern Mining, Signal Extraction
- System Integrity: Immutable State, Forensic Data Engineering, Data Lineage, White-Box Testing, 3LoD/MRM
- BI and Governance: Tableau, Power BI, Executive Dashboards, SOP Development, DMADV/DMAIC
- Domain Expertise: Systemic Risk Oversight, FCRA-Compliant Metro II Furnishing, Auto Finance, CPI, SCRA, Credit Decisioning, Expected Loss, Counteroffer Governance
“No Cold Handoffs” — engineering zero-defect, audit-ready results so stakeholders internalize the underlying “why.”
My work is built around a simple principle:
A system is not complete when the output runs.
A system is complete when the logic, evidence, assumptions, controls, and downstream interpretation are clear enough for decision-makers to trust and act on it.
That philosophy shows up across my portfolio:
- data integrity before modeling
- signal detection before remediation
- strategy testing before production change
- quantification before executive decisioning
- validation before stakeholder reliance
- documentation before handoff
All public repository artifacts use synthetic, anonymized, or demonstration data.
These projects do not expose real customer data, production credit policy, proprietary remediation rules, confidential model inputs, regulated operational pipelines, or employer-owned decisioning logic. The repositories are designed to demonstrate methodology, architecture, validation discipline, and executive communication patterns in a public portfolio setting.