About

Tarun Garg is a Junior Data Scientist at CARS24 with a passion for exploring data, extracting insights, and making predictions. He holds a Master of Science in Economics and a Bachelor of Engineering in Electrical and Electronics Engineering from Birla Institute of Technology and Science, Pilani. Tarun has 1.43 years of relevant experience in the field of Data Science. During his previous projects, Tarun has worked extensively with Natural Language Processing (NLP), Knowledge Graphs, and Artificial Neural Networks. As a Research Intern at Artificial Intelligence Institute at University of South Carolina, he conducted a study to examine if the deep learning model can compress a knowledge graph and then output the same graph with most of the semantics intact. His experiments showed that deep learning models are not able to express the full semantics of the input knowledge graph. He also worked as a Summer Intern at Walmart where he configured an architecture for Anomaly detection in power consumption, conducted Time Series analysis to check trend and seasonality, trained an Artificial Neural Network Model to predict the expected power consumption, and developed a live Interactive Dashboard for monitoring using Dash and Plotly. At CARS24, Tarun is responsible for finding possible fraud cases in car auctions, setting hypotheses, and validating them with historical data. He has reduced the cancellation rate from 18% to 15% and average loss on cancelled cars by 10k by laying various hypotheses and analyzing historical data to identify patterns causing high loss/cancellation. He has also built the Pricing Model and API for the Australia Module, scrapped data by setting up crons for various auction houses of Australia, used OCR to extract vehicle data and bid price, used vehicle condition data to predict refurb cost, and worked on clickstream and MSRP data to understand buying patterns and predict depreciation. Tarun has a strong grip over Python and has worked with various tools and technologies like Flask, AWS, XGBoost, Pandas, and MySQL. He is always eager to learn and explore new areas in Data Science.