Director

• Jul 2022 - Present

Tech Lead

• Apr 2022 - Present

Software Engineer

• Jun 2020 - Mar 2022

• Invented and applied novel algorithms for graph search, document fingerprinting, and ad clustering, co-inventor on a published patent. • Led the design and implementation of new applications and tools to improve production reliability, system usability and impact of existing techniques, improving efficiency by over 33% in some cases. • Developed Graph Neural Networks for detection of abuse in the Google Ads ecosystem, preventing over $200K in bad spend per day. • Performed large-scale infrastructure migrations, reduced tech debt burden of the team and improved latency of several systems by up to 94%. • Implemented ETL pipelines for ingesting and processing new signals to improve abuse detection, detecting ~1M fraudulent accounts with millions in bad spend. • Built a production-grade ML pipeline in TFX as well as surrounding infrastructure to detect fraud in Shopping Ads, surpassing human performance and decreasing merchant onboarding time to minutes instead of days. • Achieved an Engineering Excellence award as well as readability in C++ and TypeScript. Show less

Founder

• May 2020 - Apr 2021

• Developed a first of its kind video analytics solution for tracking performance and engagement in online learning. • Developed a first of its kind video analytics solution for tracking performance and engagement in online learning.

Trailblazer

• May 2020 - Apr 2021

• Developed a course on getting the most out of University and early-career development in Computer Science for international students. • Developed a course on getting the most out of University and early-career development in Computer Science for international students.

Teaching Assistant

• Sep 2019 - Apr 2020

• Assisted first- and second-year undergraduate students in the program by addressing queries, guiding through both programming and logic problems as a help center teaching assistant.

Undergraduate Researcher, Semantic Typology

• Sep 2019 - Dec 2019

• Performed analysis on and inference of a semantic space of dimensional adjectives through statistical techniques to cross-linguistic, parallel corpora in order to test hypotheses and explain previous observations in the area.

Undergraduate Researcher, Cloud ML

• Sep 2019 - Dec 2019

• Engineered schemes for efficient parallelization of history-based optimization methods like SAGA for Machine Learning in cloud settings, exploring several computation-communication trade-offs.

Machine Learning Engineer, Lights and Signs Team

• Feb 2019 - Nov 2019

• Worked with traditional computer vision techniques for lights and signs recognition based on CNN-detected bounding boxes. • Integrated an image augmentation pipeline for increasing neural network performance. • Worked with traditional computer vision techniques for lights and signs recognition based on CNN-detected bounding boxes. • Integrated an image augmentation pipeline for increasing neural network performance.

Software Engineering Intern, Ad Analytics

• May 2019 - Jul 2019

• Built components to perform ETL tasks, detect anomalies, check and report errors, and process automated triggers to lead to better availability and reliability of data. • Designed metrics, techniques and tooling to analyze, baseline, rank and track performance on various parameters and then generate allocation plans of a media budget of upwards of $2 million a month to assist strategy and decision making. • Created dashboards to surface findings interactively with real-time analysis for better visibility and flexibly answer business questions. Show less

Machine Learning Engineer, UAV Team

• Nov 2018 - May 2019

• Developed traditional computer vision pipeline to process drone images for the AUVSI competition using signal processing concepts to detect small, sometimes camouflaged markers in the scene. • Produced synthetically generated images for testing system robustness. • Developed traditional computer vision pipeline to process drone images for the AUVSI competition using signal processing concepts to detect small, sometimes camouflaged markers in the scene. • Produced synthetically generated images for testing system robustness.

Software Engineer, Affective Computing Group

• Nov 2018 - Feb 2019

• Trained CNN model for classifying human emotions within pictures and videos on top of face detection using Haar Cascades. • Working on improving the accuracy by growing the dataset by analyzing online content with our model and manually correcting answers. • Trained CNN model for classifying human emotions within pictures and videos on top of face detection using Haar Cascades. • Working on improving the accuracy by growing the dataset by analyzing online content with our model and manually correcting answers.

Publications Team Lead

• Dec 2017 - Dec 2018

• Designed accessible and engaging website, social media posts, posters, and other banners for the conference, reaching over 1000 users every month, and over 5000 users in the final month. • Collaborated with team leads of other portfolios and guided a team of volunteers, with deadlines, follow-ups, and regular meetings, to achieve the deliverables of publications in the conference. • Designed accessible and engaging website, social media posts, posters, and other banners for the conference, reaching over 1000 users every month, and over 5000 users in the final month. • Collaborated with team leads of other portfolios and guided a team of volunteers, with deadlines, follow-ups, and regular meetings, to achieve the deliverables of publications in the conference.

Software Engineer, Information Design Team

• Oct 2017 - Feb 2018

• Conducted user research (including interviews, surveys and events) of various student life divisions across campus to get pain points of the current ways of working, attaining insight into the true needs of the end-users. • Designed a system for consolidation, analysis and visualization of data about student life programs and services, creating an approved minimum viable product integrating with Google Apps Script to address the problem. • Conducted user research (including interviews, surveys and events) of various student life divisions across campus to get pain points of the current ways of working, attaining insight into the true needs of the end-users. • Designed a system for consolidation, analysis and visualization of data about student life programs and services, creating an approved minimum viable product integrating with Google Apps Script to address the problem.