Contributing to the AI, Data Privacy working group with OECD.AI, focusing on developing policies and guidelines at the intersection of Data Privacy and AI.Collaborated with the AI Risk & Accountability workstream to identify potential risks and establish accountability measures in AI technologies.Previously participated in the AI definition working group, contributing insights and expertise to shape the definition of an AI System (definition which was later picked up by the EU AI Act).

Helping to connect the dots between Switzerland and North America in science, education, art, and innovation

GM of the Ads Privacy and Safety team. Responsible for ensuring that the entire (Google) Ads Ecosystem is a safe and thriving environment compliant with regulations globally. Defining Google's Ads policies and building the processes and AI systems to enable their enforcement at scale. Google Ads lead on the Privacy Sandbox.

My role as a Chief Data Officer is to closely collaborate with our product, security, and legal teams to ensure that we are implementing the right technology, policies, and controls to rapidly (and safely) scale our portfolio of product to an ever-broadening audience of members and customers.  Data has been and will continue to be the lifeblood of LinkedIn.  To continue to deliver on the promise of data science, a continuous investment for excellence in data infrastructure and relevance are absolutely necessary for us to deliver on our mission and vision.  Moreover, our "Members First" approach requires us to be ever vigilant with respect to data stewardship and protection. As an Engineer, my team builds and maintains our core data infrastructure, creates and deploys AI models to personalize our members' experiences, provides analytics platforms/capabilities and analyses for our business and product portfolio. In addition, my engineering responsibilities include overseeing open source efforts and adoption at LinkedIn.In summary, my work revolves around making sure that we are able to leverage our data in ways that are effective, safe and reflective of our company values.

Leading the Data team at LinkedIn, a horizontal team building infrastructure at LinkedIn and creating personalized Member experiences through Machine Learning and AI. Making sure that our infrastructure stays steps ahead of our product demands and the requests of our AI models. Building AI models that strive to create a delightful and personalized experience for our Members. Continuing to drive our ability to make product and business decisions by leveraging our Data.

Responsible for our Data and Analytics infrastructure and products. Created and grew a team centered around building experiences and products that leverage our Data.- AI: Creating a data mining platform and ML models for personalizing our members experiences. These cover high volume realtime recommendations and ranking as well as ads targeting. Applying various information extraction techniques to standardize core data components that feed our models.- A/B testing environment: Building a decision engine on top of our core A/B infrastructure to enable rapid analysis ofexperiments. Key aspects are 1) ensuring the quality and integrity of our data throughout our data pipeline 2) uniformity of metrics definitions and computations and 3) an ability to 'simply' draw inferences about test performance across all metrics.- Online Data Infrastructure: Scaling our Data Infrastructure to support our product needs. Core components include Kafka our Open Source PubSub Messaging system, Voldemort a distributed KV storage system and our new Espresso NoSQL distributed data storage. - Offline Infrastructure: Building the tools and processes necessary to enable Hadoop to become our source of truth with respect to Data.- Search: From the Member experience to the indexers, creating a multilingual, distributed social search with a social network twist.- Social Graph: Building and deploying a new proprietary core distributed graph engine that scales with our Membership growth and the complexities of new APIs.- Open Source: Continuing to Drive LinkedIn's Open Source efforts and our involvement with the Community.

Started from a team of 2 Engineers and grew it to 100+ Engineers mostly focused on High Throughput Distributed System and Machine Learning. During these earlier dates at Linkedin, our major challenge was scaling our systems to support our exponential growth. In this time window, LinkedIn’s membership grew from 14M to about 150M and the number of page views (yearly) grew from ~1B to about ~ 27B. Started and lead 3 engineering tracks at Linkedin while at the same time building the case for Data and Machine Learning at LinkedIn. These tracks were:- Search: LinkedIn is a social network, search needs to blend your network with each request. Hence the standard cream of the crop approach won't work without serious tweaks. We built our system on top of Lucene and architected a distributed system combining real-time search with faceted navigation.- Social Graph: At Linkedin, Cloud is the service that supports all realtime inquires to the social graph. The service is core and critical all that LinkedIn does. The team had the daunting task of scaling the realtime service as well as creating new APIs to access the graph for new product feature.- Machine Learning, AI: The ML and Data Science team builds new data-driven experience for our Members. By 1) Creating and deploying new experiences such as People You May Know, Who Viewed My Profile or Jobs You May Be Interested In; 2) Building an infrastructure that allows us to manipulate Data at scale to create these experiences; 3) Online and Offline Data Mining systems for training and realtime serving recommendations and 4) Designing and building a world class A/B testing environment.In parallel to making sure we were making progress on these tracks, started two external programs to provide visibility about the quality of our work; our internal Open Source Program and top tier industry conference publications, with a focus on publishing only that which was actually deployed to the site

A member of the External Advisory Board provides invaluable industry feedback to DSI leadership on curriculum and the continuously evolving academia and industry needs. The relationship between the External Advisory Board members and DSI is a true partnership. It merges the leadership, expertise, and resources of its members with the visionary goals of the UCSB faculty in data science resulting in positive and transformative solutions to some of our world’s most pressing challenges.

Member of Search Labs. Search Labs is one of Microsoft's three labs and is mainly dedicated to Search as well as the Internet. Search Labs is headed by Rakesh Agrawal and migrated to Microsoft Research in 2007 under Harry Shum. Most of the work is really confidential but main projects where:- Creation of Galileo, a data mining platform built upon Cosmos, Microsoft's distributed computing environment.- Core ranking development covering new ranking features, feature pruning as well as all training and validation steps.- Task Based Searches as a new Search Paradigm- Personalization and vertical searches

Turn is building an automated targeted ad (mainly CPA-based) market. I essentially worked with the Chief Scientist (Dr. James Shanahan) and Dr. Jerome Friedman (Stanford). Delivered:- Metrics to evaluate the performance of predicted probabilities of actions on the live system. - Models to adjust the predicted probabilities of actions to what was observed. The model was later deployed and also used to determine ad rotation schemes.

Visiting Scholar with the department of Sociology. Had fun with Prof. Granovetter and the Silicon Valley Networks Analysis Project (SiVNAP).