-
Jul, 2020 – Present, Remote, D.C., USANeighborhood Rescue of America
Data Analytics Team Tech Lead- Scraped over 100k open crime data and shapefiles in DC using Python; automated data extraction and data preprocessing, and achieved a 30% decrease in time of data exploratory and cleaning processes; Created multi-layer interactive maps using Python folium; embedded maps on NRA website, which now serves as a tool displaying community resources for kids’ success.
- Constructed logistic regression, random forest and LGB models to predict LinkedIn responses; achieved 94% accuracy after tuning; identified feature importance using SHAP; increased 50% efficiency in notes sending and potential partner matching.
- Built a full-stack animal abuse registry web application using Django, crawled data from dynamic ebpages using Python Selenium, and extracted key features by regular expression; stored static files in AWS S3 buckets and connected clean data to PostgreSQL database in AWS RDS instance; wrote JavaScript, HTML, and CSS as frontend.
- Led team of 5 to work on various projects; adminitrated 4-week data & research internship program, coached and provided assistance to the new hires.