Research & Academic Projects

Reducing Nitrogen Loss in the Chesapeake Bay

Developed a comprehensive nitrogen supply chain model in collaboration with the U.S. Department of Agriculture (USDA), integrating 7+ datasets related to crop production, livestock, and interstate trade. The model supports market-based strategies for reducing nitrogen loss across the Chesapeake Bay region. Built interactive visualizations using Python Dash to communicate findings to stakeholders and assist in advancing USDA’s 2030 and 2050 nitrogen reduction goals.

Tools & Technologies: Python, Data Modeling, Dash, PyCharm, ETL
GitHub: View Repo

️ Interactive Dashboard – Nitrogen Loss in Food Industry

Designed and deployed an interactive public-facing dashboard to visualize county-level nitrogen loss within the Chesapeake Bay watershed. The dashboard uses Python Dash and choropleth maps to allow users to explore geographic nitrogen trends and agricultural emissions. Built as a companion tool to the USDA nitrogen reduction model and deployed on a custom domain for open access by researchers and policy analysts.

Tools & Technologies: Python Dash, Choropleth Maps, Web Deployment
GitHub: View Repo

Energy Data Collection for Renewable Plants

Processed and structured over 66,000 climate data files (wind, solar radiation, and temperature) sourced from the National Renewable Energy Laboratory (NREL), covering 11,000+ energy plant locations across the United States. Developed a climate data repository to support renewable energy forecasting and infrastructure planning. Applied K-means clustering to identify regional climate trends and automated the data pipeline using API integrations and custom ETL workflows.

Tools & Technologies: R, APIs, K-means Clustering, ETL, Web Scraping
GitHub: View Repo

Survey Modeling & Deployment – Charleston Area

Designed and deployed an R Shiny app for real-time demographic data collection. Integrated with Qualtrics and hosted on shinyapps.io to support regional social research.

Tools & Technologies: R Shiny, Qualtrics, Supabase, shinyapps.io, Data Visualization
GitHub: View Repo

EV Supply Chain Modeling

Built a simulation model to assess the life cycle environmental impact of electric vehicle (EV) production under various manufacturing and sourcing scenarios. The project evaluates the benefits of onshoring EV component manufacturing to the U.S., incorporating emissions data from global supply chains and accounting for regional grid carbon intensity. Used scenario analysis to explore trade-offs in energy use, resource efficiency, and carbon emissions across future policy pathways.

Tools & Technologies: Python, Data Modeling, Pandas, Scenario Analysis, Matplotlib

FEWsLab Lab Website – fewslab.org

Designed and developed a fully functional academic website for FEWsLab using Quarto. The site includes a team directory, project showcase, and interactive content that reflects the lab’s interdisciplinary research across food, energy, and water systems.

Tools & Technologies: Quarto, Markdown, YAML, GitHub Pages
Website: fewslab.org
GitHub: View Repo

Vehicle Crash Analysis – NYC (2012–2024)

Explored traffic crash trends and borough-wise injury distributions to support safety policy. Built dashboards and graphs in R and Quarto.

Tools & Technologies: R, ETL, Quarto, Data Visualization
Report: View Report

NYISO Electricity Forecasting

Analyzed 20+ years of NYISO electricity consumption and pricing data. Forecasted marginal prices using LSTM and detected demand anomalies using K-means clustering.

Tools & Technologies: PySpark, Time Series, LSTM, Structured Streaming
GitHub: View Repo

NYC Taxi Trip Analysis

Analyzed NYC taxi trip data to identify demand hotspots, predict trip durations, and estimate fares using classification models. Visualized ride patterns with heatmaps.

Tools & Technologies: PySpark, MLlib, Logistic Regression, DBSCAN, K-means
GitHub: View Repo

Graph Analytics on SEC Filings

Processed SEC filings using Neo4j and PySpark to map executive connections. Used graph centrality metrics to uncover patterns in financial influence.

Tools & Technologies: Neo4j, PySpark, Graph Analytics, Financial Modeling
GitHub: View Repo

Mortgage Analysis – Fannie Mae Loan Performance

Analyzed 60GB+ mortgage data from 2010–2011 to assess credit trends, FICO scores, and loan delinquency rates. Generated performance insights for quarterly reporting.

Tools & Technologies: PySpark SQL, Financial Analysis, Data Visualization
GitHub: View Repo

:::