Engineering Analyst, Trust and Safety Responsible AI Testing
Minimum qualifications:
- Bachelor's degree or equivalent practical experience.
- 7 years of experience in data analysis, including identifying trends, generating summary statistics, and drawing insights from quantitative and qualitative data.
- 7 years of experience managing projects and defining project scope, goals, and deliverables.
- 7 years of experience with one or more of the following languages: SQL, R, Python, or C++.
- 5 years of experience working in a Trust and Safety Operations, data analytics, cybersecurity, or other related environment.
- Master's degree or PhD in a quantitative or engineering field.
- Experience in designing and conducting experiments or quantitative research, in a technology or AI context.
- Experience working with Large Language Models, LLM Operations, prompt engineering, pre-training, and fine-tuning.
- Understanding of AI systems, machine learning, and their potential risks.
- Excellent problem-solving and critical thinking skills with attention to detail and the ability to think strategically and identify emerging threats and vulnerabilities.
- Excellent communication and presentation skills and the ability to influence cross-functionally at various levels.
- Drive structured and unstructured testing of novel model modalities and capabilities, collaborating with Google DeepMind.
- Lead platform and tooling development, designing engineering solutions and prompt generation strategies, leveraging Large Language Models (LLMs) to improve adversarial testing and bridge technical constraints.
- Define and ensure adherence to testing and safety standards in collaboration with cross-functional teams, including policy and engineering.
- Perform analyses, develop insights for model and product-level safety mitigations, and influence safety initiatives across Product, Engineering, Research, and Policy. Act as a key advisor to leadership on complex safety issues and represent Google's AI safety efforts externally.
- Be exposed to graphic, controversial, or upsetting content.