Portfolio

Projects

A selection of professional and personal work spanning data engineering, BI, machine learning, and NLP.

Personal · Health Equity Analysis
Phase 1 · In Progress

Social Vulnerability vs. Health Outcomes in America

A phased analysis joining CDC's Social Vulnerability Index with PLACES health data across ~73,000 US census tracts to quantify how community vulnerability predicts health outcomes.

  • Diabetes prevalence is 1.64x higher in highly-vulnerable tracts (15.6% vs 9.5%)
  • Socioeconomic vulnerability is the dominant predictor across most outcomes
  • Identified DC, Connecticut, and Louisiana as states with sharpest within-state disparities
  • Roadmap: interactive geographic maps, predictive modeling, and SHAP explainability in upcoming phases
Pythonpandasscikit-learnmatplotlibgeopandas
View Notebook → View on GitHub →
Data Engineering · Microsoft

SSAS Multidimensional → Tabular Migration

Led the migration of enterprise-grade SSAS Multidimensional Cubes to a Tabular Model, fundamentally improving the analytics infrastructure used by finance and accounting teams.

  • 200x faster experience when filtering data
  • 300% improvement in processing speed
  • Enabled independent scaling of cube components
  • Automated incident generation post-migration
SQLMDAXSSASAzure
ETL & Pipelines · Microsoft

Petabyte-Scale ETL Pipelines

Designed and maintained robust scheduled ETL pipelines processing hundreds of petabytes of raw data to deliver clean, analysis-ready datasets for financial reporting.

  • SQL and MDAX queries across massive unfiltered sources
  • Enabled financials-related decision-making for accounting staff
  • Created live production dashboards for end users
SQLMDAXAzure Data FactoryAzure Data Lake
Stakeholder Analytics · UVA

End-to-End Stakeholder Dashboards

Led individual analytics projects from intake through sign-off — meeting with stakeholders across the university to gather requirements, source raw data, and deliver tailored Tableau dashboards.

  • Owned the full project lifecycle: requirements, data, build, delivery
  • Translated non-technical needs into clear visualizations
  • Helped stakeholders onboard to the dashboards and the Tableau platform
TableauRequirements GatheringStakeholder Communication
BI Performance · UVA

Tableau Site Optimization

Contributed to the optimization of UVA's Tableau environment, with a focus on performance cleanup and telemetry-driven monitoring.

  • 20% faster workbook load times organization-wide
  • Telemetry monitoring to identify bottlenecks
  • Worked through cleanup tasks across multiple stakeholder workbooks
TableauTelemetry
App Modernization · Microsoft

Desktop-to-Web Configuration Tool Migration

Migrated a legacy configuration tool from a desktop application to a modern web-based interface, improving engineer productivity and release velocity.

  • 15% increase in engineer productivity
  • Faster and more frequent update cycle
  • Built with ASP.NET and modern web stack
ASP.NET.NET FrameworkHTMLCSS
Machine Learning · Personal

Crime Prediction Recurrent Neural Network

Investigated trends in violent crime rates across the U.S. using a Recurrent Neural Network trained on FBI historical crime data.

  • Trained RNN on multi-decade FBI crime dataset
  • Surfaced regional trend patterns and anomalies
  • Data cleaning and feature engineering in Python and R
PythonRRNNpandas
NLP & Sentiment · Personal

Newspaper COVID-19 Sentiment Analysis

Led a team project to scrape news articles from multiple sources and perform sentiment analysis focused on COVID-19 media coverage.

  • Web scraping pipeline across multiple major news outlets
  • Sentiment classification and trend analysis over time
  • NumPy and Pandas for data manipulation and modeling
PythonNumPyPandasNLP