Projects

Real-world pipelines, clinical data science, and applied machine learning

End-to-end real-world evidence pipeline using MIMIC-IV — Snowflake, dbt, propensity score matching, survival analysis, and SHAP interpretability.

Clinical Data Science

RWE Study: Vasopressor vs Fluid Resuscitation in ICU Sepsis

End-to-end RWE pipeline comparing 28-day ICU mortality outcomes using causal inference and survival analysis

RWE Snowflake dbt Python Causal Inference Survival Analysis SHAP

View Project →

Machine learning classification of EHR records as in-patient vs out-patient based on lab results and vitals.

Machine Learning

Electronic Health Records Classification

ML pipeline classifying inpatient vs outpatient admission type from lab results and vital signs

Python scikit-learn EHR Machine Learning Classification

View Project →

Designed a data product to reduce experimental cycle time for therapeutic discovery.

Data Engineering

Therapeutic Accelerator

Automated literature pipeline and SQL-backed analytics dashboard to reduce therapeutic discovery cycle time

Python PostgreSQL Power BI ETL Biotech

View Project →

Machine learning classification of malignant vs benign tumors using histological data.

Machine Learning

Breast Tumor Classification

Malignant vs benign tumor classification from histological features using ensemble methods

Python scikit-learn Machine Learning Classification

View Project →

Data narrative platform on population change and inequality in the U.S.

Data Engineering

Population Fluidity Project

Interactive data narrative on U.S. demographic change, migration, and inequality

d3.js JavaScript Data Visualization Civic Data

View Project →