I'm a fourth year Physics Ph.D. student at the William H. Miller III Department of Physics and Astronomy, Johns Hopkins University. I like to think about processes governing the interactions of the tiniest constituents of nature, and the scientific tools physicists use to perform large scale high energy experiments. My research is in collaboration with the Compact Muon Solenoid Experiment (CMS), a magnificent apparatus located at the Large Hadron Collider in CERN, Geneva.
Ten years ago, the CMS and ATLAS experiments together discovered the elusive Higgs boson, the last piece of the puzzle that “completes” the Standard Model - the framework that describes the nature of particle physics. However it has become increasingly clear that the Model is incomplete, and physicists are looking towards innovative theoretical solutions to give meaning to the remaining puzzles. I’m interested in searching for exotic processes that occur beyond the Standard Model, with a special emphasis on using techniques from Deep Learning to enhance such searches.
In my free time I enjoy taking deep dives into dark comedy dramas, listening to progressive metal and soul RnB, and doodling. I have a keen interest in good design and photography, and dabble amateurly in both.
Definitely reach out to me if anything on this page makes you curious for a chat.
Leptoquarks are unique hypothetical particles that can mediate interactions between quarks and leptons. Leptoquarks originally reared their heads as part of Grand Unification theories, where their unique properties would be useful in unifying and strong and electroweak interactions at very high energies. More recently, increasing evidence for Lepton Flavour Violation at LHCb, and recent results from the Muon g-2 experiment might point to the existence of massive leptoquarks!
As part of my thesis research at JHU, I am searching for spin-0 and spin-1 leptoquarks that mediate interactions between up/down/charm/strange quarks and electrons/muons at CMS. We diverge from the traditional bump-hunt technique because we search for our leptoquarks in Drell-Yan like events, which are non-resonant. This search channel will allow us probe leptoquarks that are much heavier than those excluded in previous searches.
The Pixel detector is the innermost part of the multi-layered multi-ton onion that is CMS. The Pixel detector tracks the movements of charged particles that are produced when billions of protons collide with each other in the LHC beam pipe. Being the innermost detector component, millions of radiation-hard silicon pixels make up the structure of the Pixel detector, providing a position resolution of $\sim 10 \mu m$, smaller than the width of a single strand of human hair.
As part of the Pixel Offline Software team, I am building a deep learning based algorithm to accurately locate particles within each silicon sensor module. Particles most often graze the modules at an angle, depositing charge in several pixels as they pass through. Convolutional Neural Networks can look at distributions of these charges and predict the true hit position of the particle with great accuracy.
I am also currently serving as the contact person between the Pixel Offline Team and the Alignment, Calibrations, Databases Team, as well the contact person between the Pixel Offline Team and the Prompt Calibration Loop Team at CMS. I coordinate various calibrations and database updates between these groups and monitor changes to the Pixel detector that are crucial to Monte Carlo production and data taking.
As part of my Master’s thesis at CPPM, Marseille in 2018, I performed a study to test the discriminatory power of Long Short Term Memory networks trained on low level raw detector features. The LSTM was applied in a search for a Higgs boson produced in association with top quarks, that decayed to a pair of bottom quarks. This channel has multiple hadrons in the final state that produce jets, making it very difficult to suppress events that produces similar final states. We also used a combination of low level features (particle momenta, angular information) and high level features (b-tagging variables, invariant masses of particles) to see if the LSTM performed better than state-of-the-art BDTs - and indeed, it did.
I have worked on several smaller projects that individually shaped my interests within CMS and more generally in high energy physics. In 2018 I briefly contributed to performance measurements of the ATLAS Pixel detector, where I investigated the effects of threshold gradients and radiation damage in the silicon sensors with the CPPM Pixel Group. For my Bachelor’s thesis at TIFR, Mumbai in 2017, I worked for the Phase-2 upgrade of the L1 Trigger at CMS, building algorithms for tau lepton reconstruction in the Electromagnetic calorimeter using FPGAs.