Pooja Jhunjhunwala
Staff Software Engineer and Tech Lead at Google
Biometrics · Computer Vision · Machine Learning Systems
Biometrics · Computer Vision · Machine Learning Systems
Hi, I’m Pooja. I’m a Staff Software Engineer and Tech Lead at Google, where I work on Pixel Biometrics. I’m particularly interested in the parts of machine learning that sit between a promising research idea and a dependable product: data and evaluation systems, real-world model behavior, on-device constraints, and the security and privacy requirements that come with technology used on personal devices.
Much of my recent work has centered on Pixel Face Unlock. As a technical lead, I have worked across its machine-learning infrastructure, evaluation, and product delivery. One milestone I am especially proud of is Pixel 8, which became the first Android phone to meet Class 3 face-authentication requirements using a conventional front-facing RGB camera rather than dedicated depth or infrared hardware. In practical terms, this meant Face Unlock could be used not only to open the phone, but also with supported banking and payment applications.
Earlier in my time at Google Research, I worked on Google Clips, Pixel Intelligent Photography, and Pixel 4’s NIR-based Face Unlock. These projects gave me the opportunity to work on a wide range of problems involving on-device computer vision, privacy, data quality, and production ML infrastructure.
My academic interests have also included natural-language processing and distributed computing. While at Texas A&M University, I co-authored CaseSummarizer, a legal-document summarization system presented at COLING 2016. The project explored how information specific to judicial opinions could be used to generate more useful summaries and presented the results through an interactive interface. It has been rewarding to see the paper continue to appear in legal-NLP research over the years.
I hold a master’s degree in computer science from Texas A&M University and have spent more than a decade working on applied machine learning and software systems.