Predicting Wheat Yields in New Environments
Compare satellite data, field records, and prediction models to investigate which wheat varieties may perform well where trials are limited.

I am available for consulting and applied research projects that require advanced technical analysis and practical experience in international development. I can help with:
If I can help with a problem you are working on, email me to discuss it. You can also download my CV and explore my past projects.
My past projects cover diverse topics, from applied development research to tools that organize and explore information in novel formats.
Compare satellite data, field records, and prediction models to investigate which wheat varieties may perform well where trials are limited.
A study workspace connecting handwritten notes, textbooks, source-linked answers, and reviewed study materials.
Find relevant papers, make their contents searchable, and look across them alongside personal research notes.
Detect patterns in how data depart from a bell-shaped distribution, and compare the proposed test with Shapiro–Wilk in small-sample simulations.
Tailor a resume from a maintained record of experience, preserving fixed details and saving checks alongside each application.
Coordinate AI-assisted work through explicit task handoffs, isolated working copies, and human-controlled integration.
My current research focuses on how communities prepare for and recover from overlapping climate disasters, armed conflict, and economic shocks. I am especially interested in whether social protection and disaster financing, such as index insurance and pre-crisis cash assistance, can protect livelihoods, reduce displacement, and improve long-term recovery. I want to examine how lessons from natural disaster response and resilience can be adapted for conflict settings, and how interventions impact communities differently in distinct contexts.
This agenda builds on my experience combining satellite observations and experimental data to compare crop-yield prediction models, applying advanced quantitative analysis and machine learning, and building reproducible tools for processing and organizing diverse datasets. I aim to combine methods from economics, climate science, machine learning and software development with my personal experience implementing development programs to better plan and evaluate humanitarian and social protection programs in rapidly changing and information scarce environments.
USAID/Ukraine
USAID/Colombia
USAID/Washington
Centre for European Economic Research (ZEW)
Columbia University
Georgia Institute of Technology
Indiana State University
University of California, Berkeley
Mercer University
The first full microeconomics concept graph makes assumptions and source references part of each connection, giving me something more useful to inspect than a cluster of related …
The first microeconomics guide connected selected sources, drafting, and reviewed edits—and showed why the finished guide still needed checking against the assignment.
Handwritten questions can lead to saved explanations, with a visible distinction between answers drawn from course materials and answers that still need sources.
Shrinking the notes history and keeping search records connected when source files move.
Building audio from mathematical notes meant translating formulas first, then discovering that a valid recording could still be far too long.
A tutor that gives hints, separates different kinds of reasoning gaps, and checks a solution with a fresh view of the problem.
How automatic drawing exports and editable plugins help me connect handwritten notes with course materials and adapt the app to my tablet.
Reusing a textbook across courses and recovering from memory failures make conversion batches easier to supervise.
Comparing abstracts with a research question and following citation trails to build a focused paper library.
Replacing a model's interpretation of a resume with a structured parser made the initial import faster and easier to check.