- Developed a quantum program scheduler akin to classical CPU schedulers, crucial for Quantinuum's H2 quantum computer launch and achieving industry-leading quantum volume 2^17, 2^18, and 2^19 milestones.

- Programmed a 78% accurate TSN AI model for action localization in untrimmed videos- Dockerized the API code of an ML algorithm library to enhance its scalability- Generated temporal action proposals using a PyTorch developed BSN AI model- Developed action localization ML/AI models for video understanding- Extracted x/y optical flow and RGB features for action recognition in untrimmed videos

- Reduced the training time of a recommendation system from 4 hours to 2 minutes- Trained 1D CNNs, LSTMs, Random Forests, and Decision Trees for time-series analysis- Trained genetic algorithms and Q-Learning models for optimizing “flow-shop” schedules- Trained collaborative filtering recommendation algorithms using the Fast-ai framework- Trained SOMs and K-means Clustering models for detecting faults in rotating machinery- Trained 2D CNNs to predict fault types in semiconductor wafer images - Used ensemble learning to predict removal rate in Chemical-Mechanical Planarization- Developed ML/AI course material for Marquette University- Performed predictive maintenance on turbofan engine data to predict its RUL

- Co-authored mathematics textbooks- Developed linear algebra, ODE & PDE courses for the University of Wisconsin System- Served on the editorial board of university textbooks in calculus, linear algebra, PDE, etc.

- Tutored courses ranging from algebra to multivariable calculus to general chemistry- Selected as the solo Sunday tutor and Diversity Center tutor