Senior Staff Engineering Analyst Manager, Gemini, Nanobanana, Veo

Google Google · Big Tech · Sunnyvale, CA +1

Senior Staff Engineering Analyst Manager for Gemini, NanoBanana, Veo, focusing on Introspection and post-training model mitigations. The role involves leading and scaling a team, pioneering safety mitigation architectures using technologies like supervised fine tuning and direct preference optimization, driving cross-functional alignment, directing foundational model launch strategies, and applying data science to audit model safety. Requires 8 years of experience in data science/ML and 5 years in technical leadership/people management.

What you'd actually do

  1. Lead and scale a team of engineering analysts focused on Introspection and post-training; develop the next wave of leaders through technical mentorship and guidance.
  2. Pioneer the use of technology (Introspection, supervised fine tuning, direct preference optimization, develop encoder systems, etc.) and design safety mitigations architectures.
  3. Drive cross-functional alignment across DeepMind and Trust and Safety stakeholders, and lead executive communications, design, and metrics reviews.
  4. Direct the strategy for foundational model launches (e.g., Gemini, GemPix, Veo, etc.).
  5. Apply statistical and data science methodologies to audit model safety and uncover vulnerabilities, ensuring continuous security enhancement through insights.

Skills

Required

  • data science
  • machine learning
  • SQL
  • Python
  • technical leadership
  • people management

Nice to have

  • prompt engineering
  • fine-tuning LLMs
  • LLMs

What the JD emphasized

  • model mitigations
  • Introspection
  • post-training
  • supervised fine tuning
  • direct preference optimization
  • safety mitigations architectures
  • foundational model launches
  • audit model safety
  • uncover vulnerabilities
  • applying machine learning/large language models (LLMs) in industry settings
  • prompt engineering
  • fine-tuning LLMs

Other signals

  • leading a team
  • developing leaders
  • technical mentorship
  • pioneering technology
  • designing safety mitigations
  • cross-functional alignment
  • strategy for foundational model launches
  • auditing model safety
  • uncovering vulnerabilities
  • applying ML/LLMs in industry
  • prompt engineering
  • fine-tuning LLMs