Engineering Manager, ML Efficiency, AI Rapid Response Team - Google

Minimum qualifications:

• Bachelor’s degree or equivalent practical experience.
• 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 with one or more of the following:
  Speech/audio (e.g., technology duplicating and responding to the human voice), reinforcement learning (e.g., sequential decision making), ML infrastructure, or specialization in another ML field.
• 5 years of experience with ML design and ML infrastructure (e.g., model deployment, model evaluation, data processing, debugging, fine tuning).
• Experience integrating generative AI tools or LLM interfaces into workflows.

• Master’s degree or PhD in Engineering, Computer Science, or a related technical field.
• 8 years of polyglot coding expertise in C++ and Python, and familiarity with Google's core infrastructure (XManager, Brain Server, Saved Model, Pathways).
• 8 years of experience with data structures and algorithms.
• 3 years of experience working in a complex, matrixed organization involving cross-functional, or cross-business projects.
• Knowledge of bridging high-velocity prototyping (TVP) with permanent distributed enterprise scale (Franchises) via structured handoff packages ("graceful exits").
• Track record leading SWAT, Pathfinding, or Forward Deployed Engineering (FDE) teams in fast-paced startup or ambiguous enterprise environments.

• Lead technical pathfinding and system design for the ML Efficiency Hub, driving complex 1–6 month Engineers and 2–4 week Strike Sprints.
• Design, prototype, and write production C++ and Python code alongside your team for model distillation, speculative decoding, dynamic batching, and distributed serving systems.
• Take ill-defined executive mandates ("The Hot Plate"), quickly de-risk technical feasibility within strict latency, FLOPs, and tokenomics thresholds, and deliver highly persuasive TVPs.
• Perform deep compute surgery on legacy P0 pipelines, evaluate complex architectural trade-offs, and establish concrete "Graceful Exit Packages" that set partner catching teams up for permanent autonomy.