Minimum qualifications:

• Bachelor's degree or equivalent practical experience.
• 1 year of experience in data management, metrics analysis, experiment design and automation.
• Experience with classification systems, ranking systems, or similar.

Preferred qualifications:

• Experience translating analytical insights into business strategies and actions.
• Experience creating opportunities to innovate, conceptualize, and complete while working with global teams.
• Experience collecting, managing, and synthesizing large data sets and information from disparate sources, statistical modeling, data mining and data analysis.

About the job:

As an Engagement Abuse Analyst, you will be a key contributor to the strategy for reducing network badness on YouTube. You will balance the short-term requirements and see the big picture to make long-term recommendations accordingly. You will be exposed to graphic, controversial, and sometimes offensive video content during team escalations in line with YouTube's community guidelines. You will work on-call at weekends on a rotational basis as needed.

Responsibilities:

• Apply advanced methods to data sets in understanding the impact of abuse to the YouTube ecosystem. Develop new workflows against known vectors of abuse.
• Learn technical concepts, systems, and deliver meaningful results using them. Communicate technical results and methods clearly.
• Maintain quality by providing feedback metrics to the Global team. Manage technological solutions for streamlining quality assurance and produce scalable training solutions to emergent workflows.
• Perform fraud and spam investigations using various data sources, identify product vulnerabilities, and drive anti-abuse experiments to prevent abuse. Work with engineers and interact cross-functionally with stakeholders to improve workflows via process improvements, automation, and anti-abuse system creation.
• Develop Machine Learning Models (e.g., using LLMs, BERT) to detect and prevent various types of abuse on YouTube. Develop Machine Learning based detection systems for different adversarial and large-scale coordinated abuse vectors.

#J-18808-Ljbffr