Liz Peng

I'm a passionate Machine Learning Scientist TL and Manager, with extensive expertise in AI/ML/LLM, Natural Language Processing, GPT prompt engineering, Classification, Recommendation systems, and end-to-end ML system design and deployment. In my most recent role over the past four years at Microsoft, I served as both Tech Lead (2 years) and Manager (2 years) of the Content Services Moderation Science team for our product - Microsoft Start (msn.com). Within Trust Safety and Intelligence, our team utilized AI/ML/LLM and GPT solutions to enhance quality, scalability, and cost efficiency. We achieved 100% automation for proactive approaches, significantly reduced human labeling costs, and delivered high moderation quality to satisfy 2000 publishers worldwide and 100 million DAUs. Prior to joining Microsoft, I spent approximately six years at Amazon, starting as a summer intern. Initially focused on developing fraud detection models for real-time transaction risk mitigation, I later diversified into various impactful projects across different teams. These included contributing to personalized recommendation models for Amazon Payment Products, leading the development of the innovative first-ever smart feature 'Did You Mean?' for home automations, to improve the interaction experience for Smart Home customers through multi-term conversations, and pioneering machine learning applications for Alexa Data services, at Alexa Core AI team.