Sharvil Katariya

Data Scientist

Microsoft•  May 2020 - August 2020

• Evaluated various model architectures to improve the Ad Click predictions for Bing’s Native Ads Team by incorporating event time into sequential models. • Improved the AUC by 8% from 88% for Ads Reranking task • Secured 2nd place in Microsoft’s C&AI Hack for Good hackathon for building a Microsoft Team’s Tab Application to provide an interactive virtual class environment leveraging Azure’s Cognitive Services as well as Bing APIs.

Senior Software Engineer

Samsung•  June 2017 - July 2019

• Built Android apps to run short text classification models that are fast & lightweight (<1MB RAM) on Samsung devices. • Implemented a JNI layer to run native code as well as improved the preprocessing pipeline on devices to run 20% faster • Developed scalable ML algorithms and production pipelines to predict smartphone users’ demographics & interests based on app usage, geolocation, browsing behavior and music/video usage as obtained on Android devices. • Built a Generic Recommendation Platform, with a plug & play architecture for the addition of any new services to generate recommendations right off the bat. • Published 3 Research Papers and filed 2 Patents as well as mentored interns.

Software Engineer

VMware•  May 2016 - July 2016

• Developed a prototype to integrate Mobile Device Management agent across multiple Smart TV platforms including Android, Tizen, FireTV & WebOS. • Created a web-based application using AngularJS, NPM, Bower & Gulp • Architected real-time processing using Apache Kafka and Spark Streaming to extract relevant information about meetings, using Outlook’s REST APIs.

State University of New York at Stony Brook

Computer Science, MS•  May 2019 - December 2020

CGPA: 3.87/4.0