资深后端开发工程师（AI计算效率方向） - Data AML
Core
Building and operating a central GPU compute pool for ByteDance's Data AML platform to support training and inference for recommendation, advertising, CV, voice, and NLP systems.
Role type
Senior IC backend engineer (AI compute efficiency)
Builds
Central GPU compute pool, AML machine learning platform, high-performance computing infrastructure
Domain
Cloud-native AI infrastructure, distributed systems, GPU optimization
Deliverable
infrastructure
Required skills
Linux, Golang, C/C++, Python, Java, distributed system design, Kubernetes, container orchestration, GPU/NPU hardware architecture, CUDA, MLOps
Preferred skills
GPU performance analysis, PyTorch, CV/NLP model algorithms, image acceleration
Responsibilities
Design and optimize resource scheduling strategies for the GPU compute pool; collect, process, and analyze GPU data for budget and efficiency management; develop and optimize the AML machine learning platform; explore cloud-native and public cloud architectures for high-performance computing.