Building production ML systems that drive business decisions. Mortality-risk models, demand forecasting, MLOps pipelines — from feature engineering to real-time inference.
I'm a Data Scientist at Prudential Financial, building mortality-risk and premium-determination models for life insurance underwriting. My work sits at the intersection of classical ML, statistical modeling, and business impact.
With 5+ years across insurance, pharma, and financial services, I specialize in taking models from Jupyter notebooks to production endpoints — with monitoring, drift detection, and automated retraining built in from day one.
End-to-end ML systems deployed in regulated, high-stakes environments.
Prudential Financial · Life Insurance Underwriting
Built an XGBoost + Logistic Regression ensemble on 800K+ policyholder records for mortality-risk scoring. Engineered 140+ features, implemented champion-challenger framework with actuarial calibration, and deployed on AWS SageMaker with real-time + batch inference endpoints.
Insurance Policy Parsing · GenAI
RAG pipeline using GPT-4 + LangChain + ChromaDB for intelligent document parsing. 512-token sliding windows, structured Pydantic validation for 40+ fields, deployed serverless on AWS Lambda with S3 triggers.
Multi-SKU Prediction · Feast
Prophet + LightGBM hybrid with shared Feast feature store serving 3 downstream models. Bayesian hyperparameter optimization on rolling cross-validation.
MLOps · Drift Detection · Airflow
PSI/CSI monitoring across 15 production models. Tiered alerting with automated retraining triggers and canary deployment with rollback on AUC degradation.
Databricks · Delta Lake · dbt
Bronze/Silver/Gold medallion on Databricks + Delta Lake. dbt transformations with Great Expectations quality checks. Power BI dashboards with row-level security serving 200+ stakeholders.
Bayesian Experimentation · Sequential Testing
Bayesian A/B testing with Thompson Sampling for dynamic traffic allocation. Sequential testing with O'Brien-Fleming spending function. Guard rails for automatic stopping on loss-ratio drift.
Prudential Financial · Newark, NJ
Building mortality-risk models and premium determination for life insurance underwriting. Champion-challenger scoring with XGBoost in AWS SageMaker. End-to-end model lifecycle from feature engineering through production monitoring.
Pfizer · New York, NY
ML models on health/demographic data for propensity and marketing measurement. Databricks + PySpark pipelines with Delta Lake and Airflow. FastAPI + Docker/K8s deployments with MLflow.
Capgemini · Jersey City, NJ
Risk assessment, pricing, and decision-support for financial services. Bronze/Silver/Gold pipelines with Spark + Delta Lake. A/B testing practices and automated Tableau/Power BI reporting.
Zensar Technologies · Pune, India
Prediction, segmentation, and measurement analyses. Clustering, PCA, and published research methods. Full model lifecycle from EDA through production monitoring.
University of Bridgeport · GPA 3.77
Open to full-time roles in Data Science, ML Engineering, and AI — remote or anywhere in the US.