Developed a web app that uses anomaly, outlier, and change detection algorithms on multiple time series data to identify global humanitarian security incidents and send danger signal alerts across 80+ countries.

Built an alternative, web-based, proprietary smart spreadsheet that handles relational data. It has multiple views for planning (Kanban view), data entry (Grid view), and data analysis (Pivot view). This saved my client $3,000+/year in subscriptions they no longer had to pay!

Developed automated webscraping tools to gather online mentions, news, and citations related to clients' research products and experts. This provided impact information to give to donors and identified addressable online and social media engagement gaps.

Used ensemble machine learning (decision trees) and PEPFAR supply chain data to develop a classification-then-regression model to predict 1 in 2 delayed HIV medicine deliveries with a reliable estimate of the number of days delayed.

Developed model to classify and predict health facility capacity then built an automated Excel/VBA-based tools for government district health managers to set targets and optimize capacity.

Investigative statistical analysis and visualization of Zimbabwe's 2018 elections using millions of rows of data. My analysis was presented as evidence in Zimbabwe's Supreme Court case review of the results.

A macro-enabled VBA Excel database and dashboard to track global commitments to funding and policy actions against SGBV. This allowed global SGBV stakeholders to visualize, manage and track funding and progress across 50+ countries and 20+ INGOs.