I'm a Cornell Ph.D. Physicist with experience in architecting end-to-end computer vision ML pipelines and automating complex metrology workflows, seeking to leverage rigorous problem-solving skills to build scalable machine learning infrastructure
Highlights of my career:
- Developed a computer vision ML solution to classify wafer map failure patterns reducing reporting cycle time by over 80%
- Accelerated electron microscope imaging throughput by 10x+ by engineering a super-resolution ML model to reconstruct high-fidelity images for critical dimension metrology, significantly reducing data acquisition time.
- Developed an ML surrogate model for electron beam simulation, achieving a 1000x speedup over the traditional process with 98% accuracy.
- Author and maintain an open-source Python project (+100 stars) for the Interactive Brokers API, supporting stock and options trading.
- Developed and deployed live market-neutral crypto strategies leveraging CFTC Commitments of Traders (CoT) data to extract sentiment-based alpha, achieving a 2.1 Sharpe ratio over the last 7 months.
- Predicted unconventional correlation between electron beam properties and laser intensity used for photoemission using Monte Carlo methods and Boltzmann equations (+200 citations).
Website: https://utilmon.github.io/
GitHub: https://github.com/utilmon
Google Scholar: https://scholar.google.com/citations?user=WmYfAsYAAAAJ