Maheshwari Public School, Ajmer

Indian Institute of Technology, Delhi

Data Scientist at Myntra

Bengaluru, Karnataka, India

Myntra

Data Scientist

- Designed and implemented causality estimation system using advance information-theoretic measures (e.g. Transfer Entropy) to facilitate root causes analysis.• Worked on time series forecasting methods (ARIMA models) and subspace (PCA) based outlier detection methods. • Implemented online version of standard seasonal and trend decomposition using Loess algorithm. • Designed and implemented dependency estimation system between nonlinear time series using multiple statistical measures ranging from pearson-correlation to maximal-information-coefficient.• Designed and implemented analytical solution for near-real time malicious URL detection using lexical features and linear support vector machines (libSVM).

Guavus

Sr. Analytics Engineer, Data Science and Machine Learning Team

- Developed Bid and Budgets optimization algorithm for online advertising campaigns on various Social and  Search networks (Facebook, Google, Yahoo, Bing)• Worked on Viral Metric Index, a single parameter to measure the virility of a Facebook Page & App.

Syncapse

Research Engineer

- Implemented distributed gaussian mixture model under MapReduce framework (Apache Hadoop)• Ideated a clustering based approach for Best-Fit-Recharge scheme for prepaid mobile subscribers.• Won the Cloud Computing Challenge worth 3,500 US$ (Amazon Web Services credits), IEEE CloudCom 2011. (Paper Published)

Flytxt

Research And Development Engineer

- Worked as a part of HP-Labs Intuitive Multimodal and Gestural Interaction (IMaGIn) Project. • Developed a real-time semi-supervised visual-speech/lip-activity detection system.

Hewlett-Packard Laboratories

Summer Intern

M.Tech in Information & Communication Technology

B.Tech in Electrical Engineering

Vikram Garg