Projects
Selected work across scientific machine learning, cybersecurity systems, and recommendation engineering.
Scientific Machine Learning / Climate Downscaling
A research software pipeline for ERA5-to-PRISM temperature downscaling, focused on high-resolution spatial reconstruction, terrain-aware residual modeling, and boundary diagnostics.
Spatial reconstruction and boundary diagnostics; historical RMSE 2.37 to 1.57
Active research codebase; temporal modeling is archived while spatial reconstruction limits are being diagnosed.
Cybersecurity / Full-Stack Systems
A Java, Spring Boot, and React research system modeling secure communication as an attack-defense-recovery workflow with adaptive security modes.
CPHS shows about 3.3x lower compromise ratio than NORMAL in the recorded high-intensity sample
Public academic/full-stack cybersecurity project; not a third-party-certified production messenger.
Recommendation Systems / Full-Stack
A local React and Node.js recommendation prototype connecting food, fitness, and media with contextual signals, feedback history, and explainable adaptive scoring.
Heuristic adaptive scoring across food, fitness, and media; no neural-model claim
Working local full-stack prototype with optional MongoDB and local JSON fallback storage.