11 hours ago
Base Salary
$174k - $252k/yr
Responsibilities
- Optimize Gemini inference throughput, accelerator utilization, batching, scheduling, KV cache efficiency, and cost per token.
- Build agentic workflows that discover, implement, and validate performance improvements at scale.
- Characterize new TPU generations and tune kernels, sharding, and serving configurations for their memory hierarchies and interconnects.
- Deliver production-ready performance improvements and provide insights for TPU hardware and compiler roadmaps.
- Implement quantization and sparsity for production Gemini models and evaluate architectures, optimizers, and training methods for reduced compute requirements.
- Create systems for agent testing using 2D and 3D games, develop test problems in physics simulators, visualize results, build agent leaderboards, and test algorithms on robots.
Requirements
- Bachelor's degree in Computer Science, Software Engineering, Mathematics, or a related STEM field, or equivalent practical experience.
- Five years of experience in software development, systems programming, or applied ML engineering, or equivalent education credit.
- Three years of experience with performance optimization, systems data analysis, visualization tools, or debugging.
- Three years of experience with Generative AI, large language models, machine learning, and related frameworks.
- One year of experience designing and launching production software products or foundation AI systems.
- Preferred: Master's or PhD in Computer Science, Artificial Intelligence, Robotics, or a related technical field.
- Preferred: Three years of experience in a technical leadership role setting architectural direction and guiding project teams.
- Preferred: Experience developing, evaluating, or serving large language models, agentic workflows, multimodal pipelines, or AI runtime APIs.
- Preferred: Expertise or strong interest in performance debugging, performance tools, compiler design, code optimization, high-performance software, concurrent programming, or multi-core computer architectures.
Benefits
- US base pay is $174000-$252000, plus a 15% bonus target, equity, and benefits.
- Benefits information is available through Google.
- The role partners with Gemini modeling, Deployment, Compiler, and TPU hardware teams in a mission-driven AI research environment.
Tech Stack
PyTorchRobot Framework
Categories
About Google
Google builds consumer and enterprise software and services including Search, Android, YouTube, Chrome, Maps, Gmail, and Google Cloud. Its business model centers on digital advertising and paid cloud, software, and hardware offerings (e.g., Pixel and Nest) for consumers, developers, and organizations. Founded in 1998 and headquartered in Mountain View, California, Google operates globally as a subsidiary of Alphabet Inc.