Senior Data Scientist, Safety & Security

Wikipedia is a trusted source of knowledge read by over 1 billion people monthly across 300+ languages, offered for free and operated independently by the non-profit Wikimedia Foundation. This role involves guiding Wikipedia’s anti-abuse and security strategy as part of the Product Analytics team.

As the Senior Data Scientist, Safety & Security, you will collaborate closely with the Product Safety and Integrity team, leveraging cutting-edge AI and machine learning techniques to address critical challenges in online safety and security. Your responsibilities include:

• Setting the strategy for measuring success in detecting bots and inauthentic account activity (e.g., sockpuppetry).
• Analyzing data to inform the development of new heuristics for detecting potentially abusive activity.
• Partnering with research scientists and security engineers to develop and evaluate AI models for identifying malicious user-authored code.

Key Responsibilities:

• Serve as a strategic partner for product managers, engineers, designers, and colleagues in product development teams.
• Proactively recommend product direction and data strategies, surfacing insights and interpreting results.
• Deliver accurate quantitative data and insights to inform strategy, guide product decisions, and assess impact in a fast-changing environment.
• Communicate results and data-backed recommendations to stakeholders across functions, ensuring understanding and guiding decision-making.
• Build and execute measurement strategies, evaluate experiments using Bayesian and Frequentist approaches, and perform quantitative research.
• Develop queries and code to work with large-scale data repositories using tools like Apache Iceberg, Hive, Druid, Presto, and Spark.
• Create automated, self-service dashboards using DBT, Airflow, and Superset to track success and health metrics.
• Help instrument features, perform data quality assurance, and facilitate A/B testing.
• Demonstrate good judgment in prioritizing workload and selecting analytical methods.

Required Skills and Experience:

• Experience with experimental design, statistics, or machine learning methods.
• Fluency in AI coding assistants for data analysis and technical tasks.
• Proficiency in R or Python and version control/command-line tools for data analysis, statistical modeling, visualization, and reporting.
• Fluency in SQL and experience working with large-scale data using tools like Hive, Presto, Druid, and Spark.
• Collaboration experience with product development teams in fast-paced environments to test, analyze, and evaluate user-facing features.

Preferred Qualities:

• Ability to explain data insights clearly to non-specialist audiences.
• Curiosity, critical thinking, and a commitment to lifelong learning.
• Empathy and discretion in handling sensitive or confidential data.
• Commitment to the Wikimedia Foundation’s mission, values, and guiding principles.
• Self-motivation and ability to navigate ambiguity and complexity.

Additional Desirable Experience:

• Prior experience in trust & safety, information security, privacy, or fraud and anti-abuse.
• Exposure to ethical data management and privacy practices.
• Contributions to Wikimedia projects or experience in other open-source projects.
• Experience with Superset, Growthbook, or similar open-source testing/data visualization tools.