Bridging machine learning, geological formation analysis, and federal energy policy advocacy for the American oil & gas industry.
View ResearchI work at the crossroads of subsurface geology, machine learning, and energy policy — translating complex technical findings into actionable insight for producers, regulators, and policymakers.
My research centers on formation prediction systems that combine gradient-boosted models with physics-based optimization to improve stratigraphic accuracy across multi-state datasets. On the policy side, I develop economic impact analyses and briefing materials for federal stakeholders at the Department of the Interior.
I believe the best technical work is work that gets used — so I care equally about rigorous methodology and clear communication.
RoboTop combines XGBoost ensemble modeling with physics-informed optimization to automate stratigraphic formation picks across heterogeneous well log datasets. The system integrates domain constraints — depositional sequences, regional dip, formation thickness priors — directly into the objective function, yielding predictions that honor both data patterns and geological plausibility. Evaluated against hand-picked formation tops across Southwest Kansas.