• AI/ML integration & automation for records

• Internal agentic tools

• Training-free stereo video generation supplemented by cycle consistency and latent manipulation

• Cross/multi-view pose estimation through vision transformer attended to trained depth maps

• Point query control for guided satellite image diffusion

VLM with RAG retrieving dense street-view vision embeddings for urban health assessment

Semi-supervised instance segmentation model for medical imaging

• Assisted in construction of Gaussian Regression model machine learning code to optimize RosettaScripts runtime

• Scripted in XML-style RosettaScripts code to simulate carefully planned packing, minimization, filtering, and scoring of protein structure models in their respective binding sites for future modeling in GROMACS and eventual protein synthesis

• Designed, built, and maintained thousands of dollars worth of computing-focused hardware infrastructure for several PhD students’ use