Google's Trust and Safety Intel Analysis team needs to deliver risk intelligence to drive decision making and systematic mitigation. They proactively identify, analyze, and prioritize trust risks, empowering Google teams and executives with actionable intelligence to make informed decisions and take timely action to mitigate threats, enhance user safety, and maintain trust in Google products.
Requirements
- 5 years of experience with SQL and Python (or similar languages) for data transformation and analysis
- 3 years of experience with foundational concepts in machine learning, including model evaluation and prompt engineering.
- Experience with metrics development and A/B testing (modeling, experimentation, and causal inference).
- Experience in data transformation, structured labeling, and creating machine-readable datasets.
- Experience in technical Trust and Safety, threat intelligence, risk analysis, or related roles, including abuse detection, adversarial tactics and model evaluations for safety and robustness.
- Familiarity with risk intelligence workflows (collection, analysis, dissemination).
Responsibilities
- Own expertise in SQL and familiarity with Python to design, build, and maintain automated data pipelines, reporting dashboards, and reusable frameworks which includes the skill to gather requirements from non-technical stakeholders and transform data into clear, actionable visualizations and reports.
- Act as a link between intelligence analysts and product, engineering, and policy teams which includes communicating with intel analysts and collaborating to ensure their outputs are integrated into product features and enforcement systems, and creating feedback loops that inform policy and system improvements.
- Develop and maintain evaluation datasets and metrics to measure the efficacy and coverage of intelligence-to-enforcement pipelines. This involves assessing the performance of automated systems in detecting and mitigating risks.
- Translate qualitative threat intelligence into technical artifacts, including structured data and machine-readable prompt sets for Large Language Models (LLMs).
- Transforming qualitative intelligence on emerging threats and abuse patterns into operational, machine-readable data, evaluations, and prompt sets that directly support product, enforcement, and policy teams.
- Operationalizing intelligence and understanding of how to structure, unstructured data for machine consumption.
- Work with qualitative analysis as you are with structured data sets and technical systems.
Other
- 5 years of experience in data analysis, including identifying trends, generating summary statistics, and drawing insights from quantitative and qualitative data.
- 5 years of experience managing projects and defining project scope, goals, and deliverables.
- Ability to navigate imperfect data and prioritize under uncertainty in a changing threat landscape.
- Excellent communication skills to collaborate with analysts, engineers, and policy stakeholders.
- Gather requirements from non-technical stakeholders
