My current research focuses on the analysis of time-series data (light curves) from ASAS-SN using machine learning techniques, with an emphasis on the detection and characterization of variable stars identified in this data. To this end, I have analyzed the light curves of over ~60 million bright stars detected by ASAS-SN and have classified these sources into variable/constant sources using a random forest classifier. The variable sources are further classified into the different variable classes using another independent random forest classifier. Through our work, we have discovered over ~220,000 new variable stars, significantly improving the completeness of these sources across the whole sky. We have made the light curves and ancillary information for ~660,000 variable stars (including the ~220,000 new discoveries) public on the ASAS-SN Variable Stars Database. The light curves of all the ~60 million sources used in this work are also public on the ASAS-SN Photometry Database. I am also leading the citizen science project “Citizen ASAS-SN“, dedicated to the classification of ASAS-SN g-band light curves.

Under the mentorship of Prof. Todd Thompson, I have recently begun exploring the synergy between wide-field photometric surveys like ASAS-SN and large-scale spectroscopic surveys like APOGEE, with the goal of identifying and characterizing compact objects (black holes/neutron stars) that are in binary orbits around stars. Thus far, we have identified one candidate compact object (likely black hole)–giant star binary (2MASS J05215658+4359220) combining APOGEE data with ASAS-SN. I also recently led the discovery of the closest known non-interacting compact object (likely black hole)–giant star binary known (V723 Mon). I am eager to continue my research in this exciting field to uncover more systems like V723 Mon!

At Cal Poly Pomona, I worked as the technical lead for the Milky Way Project, a popular citizen-science initiative on the Zooniverse platform.  Using a pipeline that I developed from scratch, I aggregated over ~3 million classifications made by ~30,000 volunteers using clustering algorithms to produce catalogues of near-infrared bubbles and bow shocks.