AI Infra Engineer · Training & InferenceAI Infra 工程师 · 训练与推理

Mission岗位使命

Build the compute and serving layer that lets our model teams iterate at research speed and ship reliable robot intelligence.构建计算与服务层，让模型团队以研究速度迭代，并可靠交付机器人智能。

Responsibilities工作职责

• Optimize distributed training, data loading, checkpointing and recovery.优化分布式训练、数据加载、检查点和恢复。
• Build low-latency inference and deployment paths for robots.建设面向机器人的低延迟推理与部署链路。
• Profile GPU systems and turn bottlenecks into measurable wins.分析 GPU 系统性能，将瓶颈转化为可度量的提升。

Qualifications任职要求

• Strong Python/C++ and Linux systems engineering skills.扎实的 Python/C++ 与 Linux 系统工程能力。
• Experience with PyTorch, CUDA, Triton, NCCL or serving stacks.有 PyTorch、CUDA、Triton、NCCL 或推理服务栈经验。
• Care about correctness, reproducibility, observability and cost.重视正确性、可复现性、可观测性和成本。

How to apply申请方式

Send a CV and a performance or systems project to hr@xvirobotics.com.请将简历和性能或系统项目发送至 hr@xvirobotics.com。