Fraud Detection • Risk Modeling • Data Engineering
I build fraud and risk systems that protect revenue, reduce losses, and support confident financial decisions.
My work sits at the intersection of machine learning, risk analytics, and data engineering.
I design end‑to‑end systems from SQL pipelines and Airflow orchestration to CatBoost fraud models deployed on AWS with a focus on reliability, explainability, and real‑world impact.
| Fraud & Risk Analytics | Machine Learning & Modeling | Data & Analytics Engineering |
|---|---|---|
Fraud Detection Risk Scoring Anomaly Detection |
CatBoost XGBoost Scikit-Learn |
Airflow (DAGs) ETL/ELT dbt |
Credit Risk (PD/LGD/EAD) |
Optuna SHAP Imbalanced Learning |
AWS (EC2/S3/Lambda) Snowflake |
Portfolio Insights |
Pandas NumPy |
Docker Git FastAPI |
- Fraud Detection Systems: Real‑time scoring engines with high‑precision CatBoost models and engineered behavioral features.
- Risk Modeling: Credit risk segmentation, anomaly detection, and model governance aligned with financial controls.
- Data Engineering: Scalable pipelines using Airflow, dbt, and Snowflake, with automated validation and monitoring.
- Model Deployment: FastAPI microservices, Dockerized inference, and MLflow experiment tracking.
- Analytics Strategy: Turning complex risk signals into clear, executive‑ready insights that support financial decisions.
- 🛡️ Fraud Detection Engine: Real‑time predictive system achieving 0.93 precision on high‑risk segments with explainable outputs.
- 📉 Credit & Portfolio Risk Analytics: Scorecard‑driven segmentation and PD modeling for financial decisioning.
- 🚲 Demand Forecasting: Cloud‑deployed forecasting API with 51% MAE reduction using TimeSeriesSplit and automated retraining.
- 🔁 Churn & Retention Modeling: Identified high‑value at‑risk customers with 0.64 precision, supporting ARR protection strategies.
Open to Fraud Analytics, Risk Modeling, and Data Engineering roles - Washington, DC