Staff ML Engineer, Search AI Generated Content Quality

Google Google · Big Tech · Mountain View, CA +1

Staff ML Engineer focused on the quality of AI-generated content (AIGC) for Google Search's proactive surfaces (Discover, Notifications). The role involves building agentic loops and evaluation frameworks to ensure content quality (factuality, freshness, coherence, tone), developing content through context engineering and feedback, evaluating quality using downstream signals, resolving system bottlenecks for low-latency serving, and driving technical innovation in AIGC quality and ML strategy.

What you'd actually do

  1. Build advanced AI-generated content (AIGC) quality frameworks and agentic flows to generate content for proactive surfaces like Discover and Notifications.
  2. Develop AI-generated content through advanced context engineering and agentic feedback loops to identify and deliver engaging stories globally.
  3. Conduct advanced quality evaluations and leverage downstream dense recommendation signals and feedback to improve the performance of the AIGC stack.
  4. Resolve key system-level bottlenecks in recommending and serving AIGC content to optimize efficiency for low-latency surfaces.
  5. Navigate high ambiguity, drive technical innovation for AIGC quality, and influence the broader organizational machine learning strategy.

Skills

Required

  • Experience building offline and online quality evaluation frameworks for Large Language Models (e.g., LLM-as-a-judge, RLHF, DPO)
  • Experience integrating generative AI tools or LLM interfaces into workflows
  • 8 years of experience in software development
  • 5 years of experience in machine learning, recommendation systems, natural language processing, or a related field

Nice to have

  • Master’s degree or PhD in Engineering, Computer Science, or a related technical field
  • 8 years of experience with data structures/algorithms
  • 3 years of experience in a technical leadership role leading project teams and setting technical direction
  • Experience optimizing machine learning inference, resolving system-level bottlenecks, and improving serving efficiency for low-latency global surfaces
  • Experience integrating large language model quality evaluation frameworks with downstream recommendation signals

What the JD emphasized

  • agentic loops
  • evaluation frameworks
  • quality control
  • factuality, freshness, coherence, and tone
  • low-latency surfaces

Other signals

  • AI-generated content (AIGC)
  • agentic loops
  • evaluation frameworks
  • factuality, freshness, coherence, and tone
  • proactive surfaces like Discover and Notifications