Led international research collaboration, developed a mathematical model with 100k nodes and 600k parameters for heart disease simulation, and accelerated it 1,000x using NVIDIA parallel computing. Created algorithms achieving 95% accuracy in disease prediction, over 95%/80% accuracy for regular/irregular disease detection, and an 8% improvement in scar detection. Published multiple first-author papers in top journals and conferences.

Traded stocks on Robinhood and TradingView, achieving 77% annual growth in 2023. Developed trading algorithms using various technical indicators. Trained machine learning models using supervised buy/sell signals as training data. Optimized annual yield by refining algorithms, feature sets, and parameters through backtesting. Coded in Python, utilizing yfinance and scikit-learn libraries.

Optimized computational speed by converting the model solver from tetrahedron-based to mesh-based. Developed a software module for non-rigid iterative closest point transformation to homogenize geometries, enabling cross-comparison.

Developed a software pipeline for data integration from major signal acquisition systems. Created adaptive graph algorithms to handle missing electrodes. Enhanced signal detection accuracy from 70% to 90%. Automated mesh defect correction and disease location detection. Implemented supervised machine learning models for electrogram classification. Applied unsupervised machine learning models for anatomical region segmentation. Designed user interfaces to support physicians during surgeries.

Led the development of a smart laryngoscope, overseeing seven iterations of prototypes. Designed 3D models and circuit boards. Showcased the product to anesthesiologists and research fellows at the Children's Hospital of Philadelphia.

Developed a mathematical model and electromagnetic control method to optimize a micro robot’s movement, reducing power consumption and improving precision. Developed machine learning algorithms to detect forgery in digital handwritten signatures. Created camera vision navigation algorithms for a quadcopter to identify targets and reach destinations accurately. Implemented a breadth-first search algorithm for a mobile robot to navigate through obstacle-filled environments efficiently.