At Stryker Global Technology Center, I build scalable AI/ML systems for medical and operating-room applications. I designed an end-to-end pipeline that converts femur bone meshes into synthetic CT scans and built synthetic 3D data generation workflows with Blender and NVIDIA Isaac Sim.
I fine-tuned and deployed a VLM using LoRA/QLoRA, PEFT, and Azure AI Studio for OT room status interpretation, while also implementing OpenGL texture mapping for dynamic visual effects. My work spans ambiguous requirements through deployment, with a focus on production-grade data and performance.
I've also built computer vision projects including railway defect detection and document verification, where a fine-tuned SegFormer reached 91% accuracy and 89% IoU while improving verification speed 3x. At WorldQuant BRAIN, I independently research and implement quantitative alphas using statistical and ML models.
