I work at the intersection of deep learning, autonomous systems, and robotics — building things that operate reliably in messy, real-world environments.
Right now I'm focused on:
- Physical robotics — control systems for physical worlds
- AI tooling — intelligent interfaces that make complex software accessible
Supplementary materials for real-time 3D image-based visual servo control for single-cell surgery.
Combines computer vision, robotics control, and medical imaging at sub-millimeter precision.
AI-powered assistant for Autodesk Revit (BIM). Natural language interface to building information models.
Full-stack: Python backend + C# Revit addin.
Collection of hands-on CV work: segmentation, object detection, generative models, and real-time systems.
ML for constrained devices — embedded systems, microcontrollers, edge inference.
Languages
ML / AI
Robotics
Infra
- RL-based exoskeleton control — policy training for lower-limb assistance
- Surgical AI — imitation learning from surgical video
- LLM applications — autonomous AI agents for research and analysis
Building at the edge of what's possible.