Argonne Scholar
I’m a computational astrophysicist who uses machine learning (ML) to tackle complex simulations, modeling, and data challenges. My work, honed through projects like my NASA FINESST-funded thesis at Georgia Tech, focuses on running and analyzing large-scale simulations as well as using ML to accelerate simulations while preserving physical accuracy. Now, as a Margaret Butler Fellow at Argonne National Lab, I’m applying these techniques to AI-driven scientific modeling, with applications ranging from astrophysics to atmospheric science. Beyond domain-specific problems, I’m driven by the potential of computational tools to extract insights from messy, real-world data to improve our knowledge and promote sustainable research and living. When I’m not optimizing code or debugging models, you’ll find me lost in fantasy novels, teaching dance, making music, or diving headfirst into a new hobby - usually after an ill-advised online shopping spree. My true bosses, however, are the four-legged housemates who demand constant tribute in food and toys.
Typically what my work week would consist of
Professional appointments & research projects
Trained production team and new engineers on videotaping & editing software, usage of professional-quality cameras, and handling of tricaster technology. Chief engineer on Wake Up Rutgers, a student-run morning news show that occurred twice weekly. Maintained fiber network and servers for Rutgers television and internet services. Communicated administratively between engineering, production, and programming teams.
Worked with Dr. Rachel Somerville and Dr. Ena Choi. Analyzed hydrodynamic simulation outputs of supermassive galaxies between redshifts 0 - 2, and determined trends in mass-to-light ratios of these galaxies. The goal of the whole project was to understand the effect of feedback from the active galactic nucleus (AGN) on star formation rates. Completed capstone honors thesis on this project and earned high honors.
Worked with Dr. Stephen Zepf and Dr. Mark Peacock. Using the Virgo Galaxy Cluster (VGC), conducted photometry as well as SDSS data munging and cleaning to obtain a catalog of ~400 globular clusters that existed in both SDSS and Hubble Space Telescope (HST) databases and were observed with the 'same' filter. Using these data, determined a gradient existed between the two telescopes' filters and determined this gradient value for HST's F475W and SDSS's F850LP filters when observing the VGC. Presented a poster for the project at the Princeton CUWiP 2017 and won a poster competition.
Tutored high school and undergrad students on various physics and math subjects as well as python coding. Had 23 students totally consistently helped students raise their grades by at least 10%.
Trained an artificial neural network (ANN) to use only galactic properties as input to determine a spectral energy distribution for simulated galaxies, replacing radiative transfer calculations. Current iteration of the ANN is applicable for IllustrisTNG galaxies and the scripts to use it have been released on github. Paper published in Monthly Notices of the Royal Astronomical Society in 2023.
Worked to create mock observations of a simulated high-redshift galaxy both with and without a central AGN to determine observational differences between a galaxy with a quiescent AGN and a galaxy without an AGN. Completed a group project exploring the contribution of central black holes to the galactic spectrum of high-redshift galaxies observed by JWST. Thesis work is creating a star formation and feedback emulator using machine learning techniques in order to accelerate large-scale hydrodynamic cosmological simulations while retaining physics fidelity.
Working as an independent researcher on machine learning applications to cosmological simulations and LLMs.
For any professional inquiries, please email me or refer to socials below.