Shrey Gupta Email & Phone Number

California, United States

Mountain View, Ca, Us

I lead the product for Ads AI team, collaborating with research scientists (across Ads, Google Deepmind), product teams, cross-functional partners (legal, policy, sales), and cross-product area leadership (across Google Ads, DeepMind, Google Research, and Core Infra). Our mission is to accelerate innovation in Ads by developing and deploying advanced AI breakthroughs.I am currently focussed on driving the GenAI/LLM initiative for the Ads Org (a critical priority for Google). In this role, I shape the the core technical and product strategy for creating and launching Large Language Models (LLMs) powered applications within the Ads organization, addressing the entire lifecycle from training/tuning strategies to input/output data provenance, legal, policy, and safety implications. Our novel applications of LLMs in Asset (Text/Image) generation, Conversational agents, and Ads ranking/quality systems (using LLMs as rater/policy agents) has resulted in over $X billion in revenue gains and strong adoption for our newer efforts.

Mountain View, Ca, Us

I led the development of the core ranking/recommendation system and metrics framework for Discover, the primary content feed for Google with over 1 Billion monthly active users. Discover is a unique, fully machine learning (ML)-driven, personalized, and adaptive feed product and in 2020, we overhauled the feed to expand beyond hard news into broad user interests which powered the next chapter of growth for Discover.My team's contributions were the biggest driver of growth for Discover, consistently drove over half of Discover's annual top-line organic growth.Key Accomplishments:* Feed Ranking Strategy: Defined and executed the product vision for ML-powered ranking, ensuring high page quality and optimizing for long-term user value. This involved leveraging and balancing diverse user signals and internal system data.* Measurement and Metric Design: Partnered with data science to establish the first unified North Star metric(s) for Discover, developed a comprehensive and novel metric framework (that established first ever tradeoff parameters for engagement and "feedbait"), and implemented survey and feedback mechanisms to enhance user experience.* AI Infrastructure: Led the transition to large-scale ML, online training, and TPU migration, significantly advancing Discover's technical capabilities.

Mountain View, Ca, Us

I led core initiatives to drastically improve the relevance and quality of advertisements for our users, while eliminating poor experiences.Key Accomplishments:* Machine Learning Quality Models: Spearheaded the development of foundational ML models that power Google's core Ads monetization engine, generating over $100+ billion in annual revenue.* Auction Marketplace Mechanisms: Designed and implemented novel marketplace mechanisms to drive quality improvements and bolster the effectiveness of our monetization engine. This led to some of biggest reduction in volume of low quality Ads on Google while simultaneously increasing revenue in long term.* AI/ML Infrastructure: Led the evolution and deployment of a robust ML infrastructure to support our quality models and marketplace mechanisms at scale. My team's work significantly improved user ad quality, resulting in a substantial and recurring annual revenue increase of over $X billion for Google.I

San Bruno, Ca, Us

As a Tech Lead on the YouTube's Creator product team, I had the opportunity to shape the emerging creator ecosystem by leading critical efforts across the stack:Examples of Key Contributions:Creator Notifications: Developed an ML-driven system to highlight the most crucial notifications for YouTube creators, resulting in significant growth in creator engagement.Creator Credits: Pioneered the 'Creator Credits' feature, facilitating the proposal, endorsement, and showcasing of credits on YouTube videos. This feature was patented and publicly launched at VidCon 2014.Traffic Attribution: Led the development of a large-scale real-time system for accurate traffic attribution, ensuring third-party verification (comScore and Nielsen) of YouTube's metrics and demographics. This has fortified YouTube's standing as a reputable mainstream advertising platform.

As a fun and personally rewarding side project, I manage online presence and search marketing for my family's small business (Cooling and heating solutions) in India. This has helped the business stay relevant and compete against online retail giants (Amazon, FlipKart) and OEMs by uniquely serving needs of the local community in the city of Chandigarh in North India.* Highest and most reviewed small business in its category in the region.* Over 60K+ views/month across search, map and ads (Google, JustDial, Facebook).* This has enabled the retail business grow 10x over 3 years and pivot from wholesale to retail.Link to business: https://bit.ly/adarshchd

Stanford, Ca, Us

Teaching Assistant for 5 graduate/undergraduate courses at the Stanford Computer Science department.* CS224N - Natural Language Processing (https://web.stanford.edu/class/archive/cs/cs224n/cs224n.1114/help.html)* CS 154 - Introduction to Automata and Complexity Theory (Prof. Jeffrey D. Ullman http://infolab.stanford.edu/~ullman/ialc/spr10/spr10.html#CONTACT)* CS 210A - Software Project Experience with Corporate Partners - Mentored 4 teams working on the projects and facilitated as a liaison between the corporate partners (BMW, Facebook, Microsoft, Samsung) and the teams.* CS103 - Mathematical Foundations of Computing* CS181 - Computers, Ethics and Public Policy (Prof. Eric Robert)

San Diego, Ca, Us

Worked on advanced 3D modelling efforts in the Corporate R&D Engineering org. One of the 6 students selected under the ' Qualcomm Technology Leaders and Innovators Program' across all the Indian Institutes of Technology (IITs).Spot me in this intern recap video :) https://www.youtube.com/watch?v=VLaEaOVwyy8

Bangalore, Karnataka, In

As a research intern, worked on and proposed a new Optimal Energy Management Policy for Energy Harvesting Sensor Nodes. The results have been published in a widely cited paper(750+ citations) in the IEEE Transactions on Wireless Communication, April 2010: https://ieeexplore.ieee.org/abstract/document/5441354

Bangalore, Karnataka, In

Research intern at the Supercomputer Education & Research Centre. The research was a part of the Summer Research Fellowship awarded by Indian Academy of Sciences.

Worked at the Centre for Advanced Machine Intelligence and Robotics.

Master Of Science - Ms, Computer Science, Artifical Intelligence And Machine Learning

Bachelor Of Technology, Computer Science And Engineering