At Mercor, I evaluate AI-generated scientific responses and analytical outputs for accuracy, reasoning quality, methodological validity, and instruction adherence.
In independent research, I designed an R-based workflow to predict soil nitrogen from NIR spectral data. I assessed its robustness with 7-fold cross-validation and Y-randomization, and interpreted selected spectral variables for chemical plausibility.
As a consultant with ESETSGlobal, I support laboratories preparing for ISO/IEC 17025 accreditation and advise production companies on quality control and quality assurance systems.
As a Research Assistant on the TETFundResearch Project, I analyze scientific datasets, interpret experimental findings, and support research design and manuscript preparation. My research interests include AI-for-science evaluation, scientific reasoning benchmarks, and reliable scientific data analysis.
