• Integrated PyTorch support into Intel’s software performance analysis tool.• Developed Python scripts and dashboards in HTML and JavaScript to analyze PyTorch, oneDNN, and oneMKL trace data.

I've been on staff at Inspirit AI since June 2022 (to June 2024). I'm proud of my high school students learning advanced AI materials.

• Built and researched on Monte Carlo models, a physics model sometimes considered to be the first generative model. • Created and owned three data preprocessing/data analysis/visualization pipelines.• Rapidly iterated in an agile workflow, leveraging Numpy, Scikit-Learn, Jupyter Notebook, Matplotlib, Seaborn and HPC. • Led the group’s collaboration to initiate its AI agenda, including cleaning and resampling time series data, metric selection, synthetic data, and Gaussian processes, working with a team of mechanical engineers and chemists.• Created a novel algorithm for a subclass of boundary value problem (1/4 of my time). Achieved instantaneous speed versus minutes through the finite element method, thus enabling real-time visualizations and its use as a function to be fitted for.

• Built the initial version of a monte carlo modeling system to solve kolmogorov equations.• Created a Python front-end to process input scripts and generate graph structures based on coordinates.