Jan 2022 – May 2023 · 1 yr 4 mos · Mountain View, California, United States · On-site

• App Ads Quality Team - Development of deep neural network based ads targeting models and back-end systems that serve app ads on Google owned and operated properties

Aug 2020 – Jan 2022 · 1 yr 5 mos · Seattle, Washington, United States · On-site

• Full-stack development of detail-page features related to size guidance and recommendation for Softlines products (apparel and shoes) on the Amazon retail website to reduce fit related returns

Mar 2020 – Jun 2020 · 3 mos

• Course: CS 33 - Computer Organization (Spring 2020)

Jan 2020 – Mar 2020 · 2 mos

• Course: CS M151B Computer Systems Architecture (Winter 2020)

Sep 2019 – Dec 2019 · 3 mos

• Course: CS M51A - Logic Design of Digital Systems (Fall 2019)
• Responsibilities includes creating problems for quizzes, home works and finals along with teaching a two-hour discussion section once every week.

Jun 2019 – Sep 2019 · 3 mos · San Francisco Bay Area

• Developed machine learning models to perform cross-modal retrieval search using images and Japanese text on the Rakuten recipe website, i.e., for a query image the model retrieves associated recipes and vice versa
• Devised a supervised approach to train the model using PyTorch where both the image and text domains are mapped to a common representation space using deep neural networks wherein the similarity of the query and data points in desired domain is measured based on a chosen metric
• Trained the model on a proprietary Japanese food data set created at Rakuten
• Best performing model achieved a median retrieval rank of 5.7 and top-10 recall of 64.9%, which is at par with the state-of-the-art on English recipe data sets of similar size

May 2017 – Jul 2017 · 2 mos · Bengaluru, Karnataka, India

• Worked on development of a web application for automating the extraction of Current Procedure Terminology (CPT) codes from a medical clinical note using LSTM-based recurrent neural networks.