Operations Strategy & Business Intelligence Analyst Intern @ IMMEC · MS Business Analytics & AI, University of Maryland
I build the layer between messy operational data and the decision someone actually has to make — pipelines, forecasts, optimization models, and dashboards that get used after the presentation ends.
I like building systems that make numbers behave.
Prince George's Coffee Site Selection — Site-selection screen over all 214 census tracts in Prince George's County, MD. Integrated Census ACS, LEHD workplace, and scraped business and review data into a common tract grain, then scored survivors in an interactive Tableau model with adjustable weights.
Python pandas Tableau Census & LEHD APIs sentiment analysis
Yelp Review Usefulness Prediction — Stacked ensemble over 1.58M labeled reviews predicting review usefulness under a hard business constraint: false positives at or below 10%. Reached 0.9081 AUC, recovering 69.1% of useful reviews at a 9.70% false-positive rate. Fourth place in a class-wide competition.
XGBoost LightGBM scikit-learn TF-IDF VADER
Uber Dynamic Pricing Decoded — Four linked Tableau dashboards explaining fare structure, surge mechanics, and geographic demand across 99,899 trips. Built with parameter-driven scenario filters so stakeholders run their own comparisons. Live on Tableau Public.
Tableau geospatial scenario modeling
Monte Carlo Marketing Budget Allocation — Simulated 58 feasible splits of a $100K monthly budget across three channels under uncertain conversion rates, click costs, and order values — ranking on probability of profit, not just expected value.
Excel @RISK Monte Carlo
Multi-Stage Supply Chain Optimization — Three-stage pipeline chaining store pricing, warehouse allocation, and vehicle routing under real capacity and feasibility constraints. $58K+ simulated daily profit.
AMPL Gurobi optimization
TransitPulse KPI Analytics — Relational model across 9+ core tables joining ridership, reliability, and safety data, with SQL views computing YoY change and crime-per-million-riders to surface multi-factor hotspot stations.
SQL relational modeling KPI design
IMMEC — Operations Strategy & Business Intelligence Analyst Intern Business intelligence, process improvement, and AI automation for operational workflows.
Roche — Healthcare Analytics Intern Replaced a manual weekly reporting cycle with SQL and Python pipelines spanning Athena, Excel, and SQL sources. Built the reporting structures behind recurring forecasting KPIs for Ocrevus and Hemlibra.
HorizonCore — Data Analyst Intern Built demand forecasting models and the dashboards on top of them, used by managers to shift budget toward higher-performing channels and to tighten inventory planning.
Executive VP of Operations, Smith Masters Student Association · Graduate Course Assistant, Government by Algorithm (AI & Public Policy)