Drimva

Google

Senior Staff Research Scientist/Engineer, Consumer Agents Post Training, DeepMind

Mountain View, CA, USA +3 · Senior Staff · Full-time · Posted Sep 28, 2026

Location
Mountain View, CA, USA
Kirkland, WA, USA
San Francisco, CA, USA
New York, NY, USA
Experience
8 years in software development; 5 years with design and architecture, and testing/launching software products. Preferred: 8 years with data structures and algorithms; 5 years in technical leadership; 3 years in a complex, matrixed organization involving cross-functional or cross-business projects.
Education
Bachelor’s degree or equivalent practical experience.

About the job

Research and develop post-training methods, datasets, and evaluation frameworks that turn foundation models into personalized, reliable consumer agents, including work on the Gemini model recipe.

What you'll do

  • Lead large-scale agent post-training research from hypothesis formulation through delivery in the Gemini model recipe.
  • Design ablation studies and build datasets and evaluation frameworks to test reasoning, reliability, and failure modes.
  • Develop research utilities for data analysis and model evaluations, and communicate findings to collaborators.

What they ask for

  • Bachelor’s degree or equivalent practical experience.
  • 8 years of experience in software development.
  • 5 years of experience with design and architecture; and testing/launching software products.
  • Experience with large language model, reinforcement learning, and deep learning.

Nice to have

  • Master’s degree or PhD in Engineering, Computer Science, or a related technical field.
  • 5 years of experience in a technical leadership role leading project teams and setting technical direction.

Pay

$262000 - $364000 USD base salary

Tools

Gemini, large language model, reinforcement learning, deep learning

AI focus

Agent post-training and personalization, including reinforcement learning, targeted data creation, persistent context architectures, and evaluation of long-horizon reasoning and agent reliability.

Model-written summary · the original posting is authoritative

First seen Oct 8, 2026 · Last verified Nov 5, 2026 (UTC)