# Abhishek Sethi > Principal Engineer at Amazon | Ex-Google Location: Santa Clara, California, United States Profile: https://flows.cv/abhisheksethi Seasoned engineering leader with wide ranging experience with Applied machine learning, Large scale distributed systems, data infrastructure and product domains such as Trust and Safety, Ads Serving / quality / optimization and consumer web services. Leveraged various forms of ML including Deep and shallow models, NLP, Multi-modal content understanding, Predictive time-series modeling for applications in abuse detection, user modeling, behavior prediction at largest scales solving critical product and user problems and driving revenue growth. Built online (sync / async) and batch infrastructure for low latency serving, Analytics, Content enforcement, ML model training / evaluation and measurement of business metrics. Built and Led high performing teams consisting of ML and backend / data infrastructure engineers. Built cross functional collaboration and led initiatives across engineering, product management, operations, policy and sales. ## Work Experience ### Principal Software Engineer @ Amazon Jan 2022 – Present | Sunnyvale, California, United States Conversation Understanding for Alexa. Developing core platform and associated language models to power multi-modal query understanding and facilitate natural conversations. ### Staff Software Engineer / Tech Lead Manager @ YouTube Jan 2020 – Jan 2022 | Mountain View, California, United States Text Content Integrity: Defending textual surfaces on YouTube (e.g. comments / chats) against adversarial abuse. We develop state of the art models to detect bad content in user generated language / dialects. and also to understand user behavior. End to end execution through policy formation, measurement, data infrastructure, training and serving detection models and enforcement infra (batch and online). ### Staff Software Engineer / Tech Lead Manager @ Google Jan 2008 – Jan 2020 Ads Integrity : Staff Software Engineer / TL Automated sensitive content classification for ad creatives across various ad formats and content types (text, image, videos, apps) across 60+ policies across all Google Ad platforms (>$100B / yr). Developed multi-modal content classifiers and related infrastructure / processes. Google Analytics : TL / TLM Founded and led development of Smart Data products for Google Analytics. This includes Smart Goals, Customer Conversion / churn probability, Smart Lists etc. Used by $XXb of ad revenue and increased revenue by $XXXm. https://adwords.googleblog.com/2015/12/use-smart-goals-powered-by-google.html GMail Ads : TL for ads quality Developed targeting, quality and optimization for Gmail native ad format, including an automated campaign optimizer (Target CPA bidding, automated targeting etc.). Increased revenue by 40% in under 2 quarters. Developed end-to-end Ads pipeline including User Modeling, predictive models for Click-through / Conversion rates predictions, Targeting, Auction, Metrics and experimentation, Budget management. Google for Work - Built domain registration and renewal flow in Apps admin console to allow customers to use Google Apps with their own domain. Integrated with external domain registrars ### Staff Engineer - Voice systems @ Earthlink Jan 2005 – Jan 2008 Responsible for development and operations of voice network components.. Technical lead for CPE device management, Voicemail and provisioning (OSS) systems. Managing a team of 4 developers performing provisioning development. Contributing role in overall service architecture, as well as Softswitch and SBC operations. Also involved in vendor relationship management. ### Senior Software Engineer @ Earthlink Jan 2004 – Jan 2005 Development of Free on-net calling product, allowing PC <-> PC voice communication. ### Software Engineer @ Earthlink Jan 2003 – Jan 2004 Development of scalable server application for email aggregation. ### Technical Yahoo! @ Yahoo! Inc. Jan 2001 – Jan 2002 Worked on Yahoo! Alerts. http://alerts.yahoo.com/. Contributed to development of distributed matching (reverse search) for keywords in News, Classifieds, Personals and numerical values for Stocks at web scale. Worked on scalability, reliability to ensure consistent operations and delivery across multiple channels (messenger, SMS, email etc.). ### Member of Technical Staff @ Stratify Inc. Jan 2000 – Jan 2001 Developed rule-based software to parse html documents to extract information like Stock quotes, document title, prices etc. Developed scripts for automated testing. ## Education ### Bachelor of Technology in Computer Science and Engineering Indian Institute of Technology, Bombay ## Contact & Social - LinkedIn: https://linkedin.com/in/abhisheksethi - Portfolio: https://www.youtube.com --- Source: https://flows.cv/abhisheksethi JSON Resume: https://flows.cv/abhisheksethi/resume.json Last updated: 2026-04-12