I lead a high-performing team of 90+ ML and software engineers in Google Ads Safety, protecting users, publishers, and Google from malicious advertising. I've hired and developed multiple senior staff AI engineers and managers. We've developed AI models that protected $Xyy billion dollars of revenue, and saved $Xyy million dollars of losses. I spearheaded Google-wide engineering efforts for scaled abuse prevention and Youtube ad scams, managing virtual teams across 10+ organizations and 300+ engineers. I've successfully hired and developed multiple ML senior staff engineering ICs and managers. My organization is comprised of 4 pillars:Prevention and Enablement: of advertisers by understanding their identity, interconnections, and developing appropriate safety-by-design capabilities.Detection and Enforcement: of fraudulent and abusive advertisers effectively, at scale, and swiftly. Foundational Modeling: techniques and tools to enable prevention and downstream detection.Foundational Intelligence: providing advanced signals, insights, and predictive threat intelligence to enable a safer ads ecosystem.

I manage 70+ machine learning and software engineers in San Francisco, Seattle and San Jose, across Airbnb's Trust Defense and Growth organizations. These organizations consist of 10 teams, with their engineering leaders/managers reporting to me. These teams work on areas such as:- Airbnb's end-to-end sign-up and login flows- User Identity Verification and Screening- User's profile and reputation- Enablement of business accounts- Risk Defenses for financial fraud, account compromise, and offline risk- Airbnb content integrity (abuse prevention in listings, messages, etc.)- User Privacy

I cofounded Trooly, a machine-learning company that provided trust solutions to some of the largest peer-to-peer companies. At Trooly, I built the founding team, set the vision and strategy of Trooly, raised funding of $10M, and grew and managed the technology team of 20 machine-learning engineers and data scientists.We developed an unprecedented solution that matches a minimal amount of input information about an individual or small business with a broad set of public and permissible digital information, in order to verify, screen and predict the trust of the entity in ~30 seconds. Our patented solution was based on a sophisticated machine-learning pipeline and information extraction and data integration technologies.In 2017, we were acquired by Airbnb Inc. Airbnb was also one of our earliest customers (since 2014).

As a Sr. Research Scientist in Google Research, I developed a system and machine-learned algorithms for extracting structured data from unstructured sources using information extraction and NLP, to enable the automated addition of triples into the Knowledge Graph.

I was part of Yahoo's Research organization, and worked on extraction and data integration techniques to enhance and deduplicate entities in their knowledge graph.