At the Yang Lab, I’m contributing to an AI-driven robotic molecular beam epitaxy platform for superconducting quantum materials and devices. A lot of my work has focused on making RHEED patterns easier for both people and algorithms to interpret during growth.

• Built a Python false-color pipeline to make subtle RHEED features more visible during process evaluation.
• Developed a computer-vision workflow for 3,000+ frame growth sequences using VGG16 feature extraction, PCA/Ward clustering, NMF, and changepoint detection.
• Used those tools to identify evolving surface states and better characterize transitions during MBE growth.

In the Hunt Lab, I’ve been working on hands-on experimental builds that combine mechanics, controls, and fabrication. It’s been a fun counterbalance to my more computational work because it forces me to think through design constraints in the real world.

• Designed and built an embedded control system with an Arduino microcontroller to drive a stepper motor with precision.
• Modeled and engineered a hydrodynamic device that generates surface waves through rotational motion.
• Used the project to explore fluid-structure interaction and wave behavior through iterative design and testing.

During my data science apprenticeship at Grainger, I worked with large behavioral datasets to understand customer patterns and support more targeted decision-making. I enjoyed the way the role combined statistical thinking, pipeline building, and business context.

• Processed and structured large multi-variable datasets in SQL and R for machine-learning analysis.
• Applied k-means clustering, PCA, and Firth logistic regression to identify behavioral patterns and predict customer outcomes.
• Built a Python and Streamlit visualization pipeline to make large datasets easier to explore and interpret.

I designed this floating wave-generating device as a compact system that could reliably produce visible, repeatable surface waves. The project combined CAD, fabrication, embedded control, and a lot of iterative trial and error. I am currently studying the interactions between two floating bodies whose diameters are comparable to the wavelength of the generated waves, allowing us to focus primarily on gravity-driven wave dynamics rather than capillary effects.

As part of the UChicago Quantum Society, I helped co-author a student-facing introduction to quantum machine learning. I wrote and presented sections on QAOA, hybrid quantum-classical algorithms, and barren plateaus, with the goal of making the subject easier for other students to approach.

This QuantumLab project brought together three connected ideas from quantum optics: Hong-Ou-Mandel two-photon interference, photon-based randomness, and polarization-encoded quantum key distribution. It was one of the projects that made experimental quantum systems feel concrete to me.

For this project, I merged and analyzed CRDC and CCD administrative datasets in R to explore how English Learner enrollment share relates to non-EL suspension rates. It was a chance to combine econometric modeling with a question grounded in education policy.