Drimva

Google

Agentic AI Software Engineering Manager, POps Engineering AI Foundation

San Jose, CA, USA +1 · Manager · Full-time · Posted Oct 13, 2026

Location
San Jose, CA, USA
New York, NY, USA
Experience
8 years managing and developing managers; 8 years designing software architecture, including AI/ML systems; 8 years building and architecting ML systems and infrastructure; 5 years in people management, supervision, or team leadership. Preferred: 15 years of software engineering management experience; 5 years applying advanced AI methods.
Education
Bachelor’s degree or equivalent practical experience.

About the job

Leads a team building agentic AI systems and infrastructure for Google’s workforce lifecycle, from applied AI research through production deployment.

What you'll do

  • Set technical direction and execution for agentic AI architectures, including methods to improve accuracy and reduce hallucinations.
  • Lead, mentor, and develop AI/ML engineers working on algorithms, model tuning, alignment, and agent deployment.
  • Oversee evaluation, serving infrastructure, reliability improvements, and cross-functional work to address productivity and quality issues.

What they ask for

  • Bachelor’s degree or equivalent practical experience.
  • 8 years of experience managing and developing managers.
  • 8 years of experience designing software architecture for applications, including AI/ML systems.
  • 8 years building and architecting ML systems and infrastructure; 5 years in people management, supervision, or team leadership.

Nice to have

  • 15 years of software engineering management experience building and managing technical engineering teams.
  • Experience with advanced AI methods, fine-tuning, reinforcement learning, reward models, classifiers, or agentic workflows.

Pay

$262000 - $364000 USD base salary

Tools

Reinforcement Learning (RL), Reward Models (RM), LoRA/SFT

AI focus

Leads development and production deployment of agentic AI, applying advanced AI/LLM research, reinforcement learning, reward models, classifiers, fine-tuning, alignment, evaluation, and serving infrastructure.

Model-written summary · the original posting is authoritative

First seen Oct 21, 2026 · Last verified Nov 4, 2026 (UTC)