Customer Engineer, High Performance Computing (English)

About the job

When leading companies choose Google Cloud, it's a huge win for spreading the power of cloud computing globally. Once educational institutions, government agencies, and other businesses sign on to use Google Cloud products, you come in to facilitate making their work more productive, mobile, and collaborative. You listen and deliver what is most helpful for the customer. You assist fellow sales Googlers by problem-solving key technical issues for our customers. You liaise with the product marketing management and engineering teams to stay on top of industry trends and devise enhancements to Google Cloud products.

Specialist Customer Engineers seek to accelerate Google Cloud customer success by providing domain-specific technical expertise, while responsibly influencing products and solutions.

You will partner with technical Sales teams as a subject matter expert in High Performance Computing (HPC) to differentiate Google Cloud to our customers. You will help prospective and existing customers and partners understand the power of Google Cloud, develop creative cloud solutions and architectures to solve their business challenges, engage in proofs of concepts, and troubleshoot any technical questions and roadblocks. You will use your expertise and presentation skills to engage with customers to understand their business and technical requirements, and present practical and useful solutions on Google Cloud. You will have excellent technical, communication and organizational skills.

You will be a part of a team of fellow Googlers working in an environment of respect where we promote equal opportunities to succeed.

Responsibilities

• Drive pre-sales technical strategy and build go-to-market solutions with customers and data center hosting partners to accelerate business cycles, resolve technical objections, and land large-scale Google Cloud TPU agreements and deployments across APAC.
• Serve as a trusted technical advisor on end-to-end architecture and integration strategies across the TPU software stack and customer or partner environments (fleet management systems, job schedulers, IaaS operations tooling, ancillary storage/compute, and data center networking/WAN).
• Guide customers and hosting partners through technical qualification, pod layout design, system turn-up, and operational readiness-covering slice lifecycle management, OS and firmware upgrades, telemetry integration, and hardware maintenance workflows.
• Conduct workload performance profiling, workload migration, and optimization of large-scale training and inference models and data pipelines specifically for Google Cloud TPU hardware topologies (e.g., 3D Torus over Inter-Chip Interconnect and Optical Circuit Switches).

Qualifications

Minimum qualifications:

• Bachelor's degree or equivalent practical experience.
• 10 years of experience with cloud-native or high-performance computing (HPC) infrastructure architecture in a customer-facing pre-sales, solutions architecture, or deployment engineering role.
• Experience developing, deploying, and orchestrating distributed AI/ML workloads using modern frameworks (e.g., PyTorch, JAX, vLLM, or SGLang) and job schedulers (e.g., Kubernetes or Slurm) on AI accelerator hardware.
• Experience with Linux host software, cluster control planes, and data center infrastructure integration (compute, high-speed networking, and storage).
• Ability to communicate in English fluently to collaborate with regional customers, data center partners, and global engineering teams.

Preferred qualifications:

• Experience integrating large-scale accelerator clusters with customer and data center partner software stacks, including fleet management APIs, IaaS operations tools, out-of-band hardware management (BMC/Redfish), and Prometheus telemetry pipelines.
• Experience designing, turning up, and operating large-scale AI/HPC infrastructure farms, including data center space, power, liquid cooling (CDU), Optical Circuit Switches (OCS), Top-of-Rack (ToR) networking, and hardware RMA/repair playbooks.
• Experience with lower-level performance profiling and kernel optimization tools (e.g., XProf, PyTorch Profiler, Pallas, or MaxText) to diagnose and resolve distributed training and inference bottlenecks.
• Ability to communicate in Mandarin fluently to support customers in the region.