Software Engineer @ Microsoft · MS in CS @ Northwestern University
I build software at scale and go deep on foundation models, VLMs, and AI agents. Previously worked on computer vision, materials science AI, and generative models in grad school.
I write detailed research paper breakdowns on Substack — the math, the architecture, the tensor shapes.
Selected work:
- LLM_Atom_Gen — Generating nonexistent atoms with Large Language Models
- SemiconCLIP — CLIP-based atom matching in semiconductor microscopy
- CrossPropertyTL — Cross-property deep transfer learning
- Classifier-GAN — GAN experiments for discriminator accuracy
- EBSD_FFT — FFT-preprocessed neural networks for EBSD data
- Efficient-Weight-Initializer — Weight init strategies to capture data structure before training