- developed perception module by applying deep learning models for object detection- implemented neural network regression approaches using an end-to-end Transformer architecture for lane detection

applied state-of-the-art deep learning techniques such as convolutional neural network, residual neural network and attention-based transformer to classify gamma and pi-0 jets in relativistic heavy ion collisions

The ALICE Experiment is recording around 10 PBs every year. This amount of data is stored on different storage elements all over the world. To analyze a single dataset, on the order of 10^5 files have to be read where those files containing data on the order of 50 TB. In the ALICE experiment hundreds of scientists are analyzing big datasets on a Grid system. High throughput and short turn-around times are achieved by a centralized computing system called the LEGO trains. As a LEGO train operator, I operate the LEGO train runs for analyses by ALICE scientists, and maintain the datasets in the LEGO train system.

As a member of the ALICE collaboration at the Large Hadron Collider, I contribute analyses of identified hadron correlations to advance the understanding of particle productions and the properties of the matter produced in relativistic heavy-ion collisions.

As one of the three ALICE Junior Committee Members, I coordinate ALICE junior activities and issues such as Junior Representative Election and Juniors Day at ALICE Weeks, for a community of about 500 young professionals including students, postdocs and scientists.

As the Shift Leader of the ALICE Experiment, I coordinate the activities of the shift crew in the ALICE Run Control Center during Pb-Pb and pp collision data taking. In particular, only the most experienced shift leaders get selected as the Shift Leader for the critical Pb-Pb data taking.