United States
Currently working on YouTube's personalization infrastructure, the stack for serving petabytes of data at hundreds of millions of QPS powering YouTube recs, GenAI workflows and LLMs, and deep learning. I also work on the bleeding edge of an intersection between GenAI, reinforcement learning, Ads prediction + modeling, and deep neural nets. I work with partners across Alphabet - Shopping, Gemini, YouTube Ads, Search Ads, YouTube Organic in developing SOTA models for powering the next generation of content targeting across YouTube. I've built large-scale data and telemetry pipelines for Azure machine learning services. Prior to that I worked across teams to integrate Bing search index with an Azure app streaming service, enabling an in-browser experience. I have a background in distributed systems. I have worked with talented people on many teams, shipped projects, and am always on the lookout for interesting opportunities.
Tech lead for the YouTube Discovery Infrastructure team: - Working on the bleeding edge of the world's largest training data, handling petabytes of data at scale, and interfacing with GenAI, reinforcement learning, and deep learning teams across YouTube and Alphabet. - Improve reliability and freshness of data served to Discovery systems across Google & YouTube (ranking, filtering, scoring, GenAI, profiles, deep learning). - Evaluate new technologies and drive the future technology roadmap for delivering personalized content to billions of users around the world.
I'm the TL for all YT user data used in ML training & inference powering recommendations. We build and manage infrastructure handling user data for billions of users on YouTube. This includes being up to date with Federal / State privacy laws, EU regulations, running infrastructure efficiently, lean, and in a manner that continually supports innovation for client teams while keeping user data safe.
Part of the infrastructure team for Google Nest Store. - Designed and built services using GCP to handle order flows, subscriptions, purchases and invoicing. - Worked on end-end pipelines used in surfacing order data for analytics, inventorying as well as for business visibility.
(2016 Software Engineer) Developed and launched a new auth and auth system for SSO using Azure Active Directory (AAD), integrating Intune AAD groups. Building a new PaaS service to aggregate GPS information from devices, to manage corporate enrolled devices using fence related policies (GPS and Time Fence) (2015 Software Engineer) Cross-product analytics and telemetry pipeline using Azure Service Fabric providing near real time storage and data retrieval capabilities Front-end for the Azure RemoteApp SaaS service enabling a desktop in the browser experience. Integrated the RemoteApp service with Bing to include content from Windows and Android apps into Bing's indexing service
I was involved in leading class discussions, holding office hours every week and taking an active part in almost all aspects of the course for a class size of about 200 students. Getting to interact with undergraduates and re-visiting algorithms were the highlights of this experience.
RDP iOS Team !