Head Of Product Data Science, Trust And Safety
Current Established a new data science team structure, clarified roles & responsibilities, and composed team mission, vision, and strategy. Created essential team documentation such as data science team introduction, project alignment & scope, roadmaps, bimonthly OKRs, biweekly KPI metrics review, onboarding guides, hiring SOPs, and updated job family leveling and core competencies to accurately reflect the roles and responsibilities of a TikTok Trust and Safety Data Scientist. Pioneered Project Texas as Lead Data Program Manager at TikTok, overseeing the strategic migration of dashboard/data assets and tool functionalities to enhance business interoperability, ensure data quality, and support decision-making processes, significantly reducing timelines and facilitating the launch of 13 key features. Established and directed a red team dedicated to ethical hacking and vulnerability assessments of AI and ML systems, particularly large language models (LLM’s), to preemptively identify and ensuring all data science projects adhered to principles of fairness, accountability, and transparency. Collaborated with ML engineers to enhance and develop content moderation models, achieving a 35% improvement in detection accuracy and reducing false positives by 25% Conceptualized and implemented a comprehensive metric framework, aligning KPIs across executive, departmental, and team levels to streamline data-driven decision-making; significantly enhanced organizational alignment and performance monitoring Collaborated with the Product team to establish biweekly KPI metrics, enhancing account segmentation and visibility into key issues such as overkill and appeals by categorizing accounts based on follower count. Established a robust experimentation framework for advanced ML models and Trust and Safety feature, enhancing product safety and user experience
