Staff Software Engineer, User AI Flywheel, Search Intelligence

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

Staff Software Engineer focused on the User AI Flywheel for Google Search, aiming to enhance LLM performance, quality, and capabilities through user signals and advanced modeling techniques. The role involves innovating on LLM improvements, partnering with research teams, and integrating LLM technologies into products. Responsibilities include advocating for modeling/tuning/optimization, designing user signal-driven improvements, collaborating with research, and demonstrating expertise in system design, ML modeling, and coding for LLMs. The role requires experience in software development, ML design, ML infrastructure optimization, and GenAI techniques, with preferred experience in RLHF, fine-tuning, and ML systems production.

What you'd actually do

  1. Advocate the development of modeling, tuning and optimization techniques to enhance Large Language Model (LLM) performance, quality, and capabilities.
  2. Innovate in the design and implementation of user signals driven improvement of LLMs, to address known loss patterns and response quality behaviors and capabilities.
  3. Partner with experts from Google Research and DeepMind to advance LLM technologies and integrate them into groundbreaking products.
  4. Demonstrate exceptional teamwork, sharing your expertise in system design, machine learning modeling, and coding practices tailored to the LLM domain.
  5. Cultivate strong relationships with stakeholders across the organization to collaboratively define and execute LLM initiatives.

Skills

Required

  • software development
  • software design
  • software architecture
  • ML design
  • ML infrastructure optimization
  • model deployment
  • model evaluation
  • data processing
  • debugging
  • fine tuning
  • GenAI techniques
  • LLMs
  • Multimodal
  • Large Vision Models
  • language modeling
  • computer vision

Nice to have

  • reinforcement learning with human feedback (RLHF)
  • LLM modeling
  • data collection
  • evaluation
  • fine-tuning
  • optimizing large language models
  • ML systems
  • production infrastructure

What the JD emphasized

  • 8 years of experience in software development
  • 5 years of experience testing, and launching software products, and 3 years of experience with software design and architecture
  • 5 years of experience leading ML design and optimizing ML infrastructure (e.g., model deployment, model evaluation, data processing, debugging, fine tuning)
  • 2 years of experience with GenAI techniques (e.g., LLMs, Multimodal, Large Vision Models) or with GenAI-related concepts (language modeling, computer vision)
  • Experience in reinforcement learning with human feedback (RLHF) as applied to LLM modeling, data collection, and evaluation.
  • Experience in fine-tuning and optimizing large language models, demonstrating a deep understanding of performance enhancement techniques.

Other signals

  • LLM performance
  • LLM quality
  • LLM capabilities
  • user signals
  • loss patterns
  • response quality
  • LLM technologies
  • LLM domain
  • LLM initiatives
  • GenAI techniques
  • LLMs
  • Multimodal
  • Large Vision Models
  • language modeling
  • computer vision
  • reinforcement learning with human feedback (RLHF)
  • LLM modeling
  • data collection
  • evaluation
  • fine-tuning
  • optimizing large language models
  • ML systems
  • production infrastructure