SENIOR APPLIED SCIENTIST · DANDY
I work on multimodal generative models for the 3D design of bespoke prosthetics at Dandy. These models require precision at the scale of tens of microns, and feed directly into complex manufacturing lines with little or no human intervention. Previously, I led applied machine learning work at Celonis and held research roles at Microsoft Research, J.P. Morgan, and Spotify.
01 / SELECTED WORK
Selected problems, approaches, and outcomes from my work in applied machine learning.
Dandy / Senior Applied Scientist
I’m the technical lead for a roughly five-person team taking prosthetic-design automation from concept to production, generating thousands of designs daily. My work spans diffusion and flow-matching models, crown refinement, geometric validity checks, and point-cloud transformers for anatomical segmentation.
Celonis / Senior Applied Scientist
I developed custom process-event tokenization and fine-tuned LLMs for event prediction. I also led a roughly five-person team taking enterprise LLM and agent applications from research to production: financial-data access, document-flow tracing, and live dashboard creation, deployed across dozens of globally distributed servers.
Related publicationsEARLIER EXPERIENCE
Researched neural and graph-based methods for fileless attack detection, including a weakly supervised model trained on billions of shell commands.
Research InternDeveloped computational methods for stochastic differential equations and options pricing, improving the accuracy and efficiency of long-term forecasts.
Quantitative Research Summer AssociateDeveloped podcast recommendations from sequential listening patterns and an unsupervised system to detect platform misuse from consumption data alone.
Research InternNew York University / Ph.D. in Computer Science
My doctoral work, advised by Andrew Gordon Wilson, studied how to give Gaussian processes and neural networks useful structure: uncertainty over kernels, learned invariances, and more diverse model ensembles. Those ideas became papers at NeurIPS and ICML and still shape how I approach applied research.
02 / PUBLICATIONS
Work from my Ph.D. and industry research spans process intelligence, uncertainty, forecasting, and the structure of deep learning models.
Browse publications Google Scholar03 / ABOUT
I earned my Ph.D. in computer science at NYU and a B.S. and M.S. in applied mathematics at the University of Colorado. I’m drawn to problems where strong research has to survive contact with complex data and real world constraints.
Outside work, I enjoy reading, coffee, weightlifting, and running. I’ve also completed two 90-mile endurance canoe races.
Download résumé
