Software Engineer, Commerce Content Safety

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

Software Engineer role focused on building and deploying ML models, including LLMs, for content safety in Google Commerce. This involves designing, building, and iterating on models for policy violation detection across text, image, and video, as well as developing pipelines for training data curation and model evaluation. The role requires collaboration with various teams to integrate safety measures and analyze data to improve systems. Experience in AI/ML is required, with a preference for generative AI and AI algorithms.

What you'd actually do

  1. Design, build, deploy, and iterate on machine learning models (including LLMs) to detect policy violations in Shopping content for various moderation tasks, including text, image, and potentially video analysis.
  2. Develop and deploy scalable pipelines for high-quality training data curation, striving for gold-standard datasets and implementing frameworks for model evaluation and performance tracking.
  3. Collaborate closely with product managers, data analysts, operations teams, and other engineering teams to understand requirements, translate them into technical solutions, and deliver impactful results.
  4. Collaborate with other engineering teams (Search, Gemini, Shopping Infra, Compliance, Merchant Tools, Ads safety) to integrate and support content safety measures.
  5. Analyze data to identify patterns of abuse, measure the effectiveness of our systems, and guide future improvements.

Skills

Required

  • Python
  • C++
  • artificial intelligence
  • machine learning

Nice to have

  • generative AI
  • AI algorithms
  • data structures
  • algorithms
  • LLMs
  • text analysis
  • image analysis
  • video analysis

What the JD emphasized

  • machine learning models
  • LLMs
  • content safety
  • policy violations
  • training data curation
  • model evaluation

Other signals

  • LLMs
  • content safety
  • policy violations
  • training data curation
  • model evaluation