I build the tools that turn spatial data into decisions.

My work starts from a simple observation: the most useful spatial analysis rarely stays inside a GIS. It ends up in a dashboard a policymaker checks weekly, a model a farmer trusts before planting, or an app a field technician opens with muddy hands. So I've spent my career learning both halves of that problem — the science that makes an analysis correct, and the engineering that makes it usable.

I teach geospatial science and geovisualization at the University of Tennessee, Knoxville, where my research centers on hydroclimatology — concurrent drought-pluvial events, precipitation reanalysis, and predictive time-series modeling across the Southeastern United States. That research runs on the same statistical and machine-learning toolkit I use in industry work: XGBoost, Bayesian methods, fuzzy time-series and ANFIS models, spatial regression, and increasingly, LLM-backed research tooling.

Outside the classroom, I design and build software as a founder — most recently the full stack behind an environmental-technology venture, built from the ground up: a GPS-locked field data collection platform with offline-first sync and AI-assisted photo analysis, built for technicians working with no signal in the field, plus the commercial web application around it. I've also built a cross-platform desktop GIS application distributed through an automated CI/CD pipeline, a spatial market-intelligence scoring engine used for site-selection analysis, and full-stack web applications spanning e-commerce, real-time multiplayer systems, and AI-assisted tooling.

The throughline across all of it is the same: take a spatial or environmental question seriously as science, then build something a real person will actually use.