Program Manager III, Software Engineering, YouTube

Google•San Bruno, CA

10h•$159,000 - $230,000

About The Position

The Program Manager will drive YouTube’s multi-year ML infrastructure strategy across a portfolio spanning training modernization, next-generation serving, and frontier model scaling. The role addresses compute and architectural bottlenecks for frontier AI models and helps shape how engineering and research teams build recommendation systems used by over two billion people daily. The position provides an end-to-end view across YouTube’s AI infrastructure stack, coordinating the evolution of modeling techniques, distributed training frameworks, and low-latency production-serving platforms. The Program Manager connects recommendation research scientists, systems software engineers, Site Reliability Engineers, hardware operations, and Google-wide platform teams; leads infrastructure transitions; establishes operational excellence for compute allocation and duty cycles; and partners with engineering leadership on YouTube’s AI infrastructure direction.

Requirements

Bachelor's degree or equivalent practical experience.
5 years of experience in program or project management.
Experience with the software development life cycle and managing engineering programs.
Experience working with engineering teams.

Nice To Haves

5 years of experience managing cross-functional or cross-team projects.
Proven experience driving large-scale software migrations or platform modernizations across dozens of participating engineering teams without regressing product latency or stability.
Technical background, such as a degree in Computer Science, Engineering, or Mathematics, or experience with software engineering, scripting, distributed systems architecture, or deep learning systems.
Working AI fluency and familiarity with machine learning concepts, including model training, inference serving, evaluation, and compute accelerators, with the ability to converse credibly with ML engineers and researchers.
Familiarity with compute resource planning, capacity budgeting, hardware efficiency, or accelerator duty-cycle analysis (TPUs, GPUs).

Responsibilities

Lead cross-team migrations to next-generation ML frameworks, runtimes, and training systems across distributed models.
Drive the strategic roadmap and operational delivery for production ML serving platforms, ensuring low-latency inference, high availability, and zero-downtime rollouts.
Partner with capacity planning and engineering teams to optimize compute scalability, hardware efficiency, and accelerator (TPU/GPU) utilization.
Manage cross-functional dependencies, milestones, and technical blockers across modeling, systems, SRE, and platform teams.
Establish operational cadences, proactively manage technical risks, and communicate program status, roadmaps, and trade-offs to executive leadership.
Lead complex, multidisciplinary projects from start to finish, including planning requirements, managing schedules, identifying risks, and communicating with cross-functional partners.