The AI Safety Protections team uses cutting-edge, AI/LLM solutions to develop the next generation of foundational models and ensuring safety alignment, including Gemini, NanoBanana, Veo, and Robotics. We are a team of dedicated data science and machine learning experts focusing on:
- Developing introspection heads/classifiers embedded within the Gemini transformer architecture (or diffusion latents) to steer model responses to safe directions.
- Post-training for foundational models (Gemini, GemPix, and Veo). This includes Gemini Domains (e.g., Health, Mental Health, Financial Advice), frontier safety, elections, as well as post-training for GemPix and Veo text station, image station and video station.
- Building LLM-based safeguards (Grail) to protect model outputs.
- Generating log-based insights and defining critical model metrics.
In this role, you will apply the latest advancements in AI, collaborating with experts and leadership across Google DeepMind, Core, and Trust and Safety to develop state-of-the-art model mitigations that protect users worldwide.
The US base salary range for this full-time position is $189,000-$274,000 + bonus + equity + benefits. Our salary ranges are determined by role, level, and location. Within the range, individual pay is determined by work location and additional factors, including job-related skills, experience, and relevant education or training. Your recruiter can share more about the specific salary range for your preferred location during the hiring process.
Please note that the compensation details listed in US role postings reflect the base salary only, and do not include bonus, equity, or benefits. Learn more about
benefits at Google.
Responsibilities
- Lead and scale a team of engineering analysts focused on Veo and Robotics; develop the next wave of leaders through technical mentorship and strategic guidance.
- Pioneer the use of advanced technology (Introspection, supervised fine tuning, direct preference optimization, develop encoder systems etc.) and build Google's next generation foundational models.
- Drive cross-functional alignment across Google DeepMind (GDM), Trust & Safety stakeholders, and lead executive communications, design and metrics reviews.
- Direct the strategy for high-stakes foundational model launches (Veo, Robotics, GenieVeo etc.).
- Apply rigorous statistical and data science methodologies to audit model safety and uncover vulnerabilities, ensure continuous security enhancement through actionable insights.
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.
- 5 years of experience managing teams with performance management, mentorship, and coaching of team members.
Preferred qualifications:
- Master’s or Ph.D. degree in a quantitative discipline (e.g. Computer Science, Statistics, Mathematics, Operations Research etc.).
- Experience with prompt engineering and fine-tuning LLMs.
- Excellent written, verbal, and presentation skills to effectively communicate with a variety of stakeholders, including executive leadership.
- Excellent problem-solving and critical thinking skills with attention to detail in an ever-changing environment.