Software Engineer Iii, Ai/ml, Discover AI Feed

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

Software Engineer III, AI/ML role focused on designing and iterating on agentic flows for user intent extraction, query generation, content retrieval and synthesis, and multi-modal generation within Google's Discover personalized content recommendation feed. The role involves conducting applied research, building quality intuitions, defining evaluation metrics, integrating new content formats, and implementing ML solutions at production scale.

What you'd actually do

  1. Design and iterate on agentic flows for user intent extraction, query generation, content retrieval and synthesis, and multi-modal generation. Conduct applied research on novel techniques for the above, and come up with proper solutions at production scale.
  2. Build quality intuitions by analyzing user behavior and product needs. Define the evaluation metrics and measurement approach. Iterate on quality via offline evals and live experiments to drive end user impact.
  3. Integrate the new content format with the Discover indexing and serving system; experiment with different retrieval, ranking and user targeting approaches to optimize user experience.
  4. Collaborate across teams and work with a cross-function group (Team Leads, Engineers, Product Managers, Data Scientists, Partner Teams) in Discover.
  5. Implement solutions in one or more specialized ML areas, utilize ML infrastructure, and contribute to model optimization and data processing.

Skills

Required

  • software development in C++
  • Speech/audio technology
  • reinforcement learning
  • ML infrastructure
  • model deployment
  • model evaluation
  • optimization
  • data processing
  • debugging

Nice to have

  • machine learning algorithms and tools
  • TensorFlow
  • Deep Learning
  • Natural Language Processing
  • large language models applications
  • agentic flows
  • user modeling
  • recommender systems
  • personalization
  • statistical methods
  • software development
  • excellent mathematical skills

What the JD emphasized

  • agentic flows
  • multi-modal generation
  • applied research
  • production scale
  • evaluation metrics
  • offline evals
  • live experiments
  • indexing and serving system
  • retrieval, ranking and user targeting approaches
  • ML infrastructure
  • model optimization
  • data processing

Other signals

  • design and iterate on agentic flows
  • conduct applied research on novel techniques
  • define the evaluation metrics and measurement approach
  • integrate the new content format with the Discover indexing and serving system
  • implement solutions in one or more specialized ML areas