- Intern on the Commerce Growth Intelligence team in Marketplace Org- Created a standardized listing filtering system used in production for all growth channels with 9 Integrity and Quality filters. Implemented 2 modularized APIs for Marketplace Push Notifications and Newsfeed Showcase.- Worked with Badging Models to improve the Click through rate when user gets notifications on homepage. Enabled recurring training, migrated to a new MTML model, performed score tuning, and deprecated legacy models.- As part of a stretch goal adopted the Engine-Rule Architecture for different team products to improve system quality.- Key metrics: ~6,500 significant lines of code contributed, 52 code pushes, 10+ documentation docs.

Working with Professor Rahnev Dobromir in the Computational Cognitive Science Lab. Developing and training regression artificial neural networks (ANNs) for precise quantity estimation tasks to enhance accuracy in perceptual data processing.Using Python frameworks like Tensorflow, Keras and Matplotlib for visualization.

Worked with the Automated Algorithm Design Project under the Stocks subteam. Applied deep learning and genetic programming principles to obtain more accurate results for the midpoint price prediction problem in high-frequency stock trading markets.Used Python libraries like Sckit-Learn, TensorFlow, Pandas etc. Worked on Georgia Tech’s Supercomputing cluster PACE-ICE to deal with heavy datasets.