AI Infra技术负责人（创作大规模训练方向） - 剪映CapCut
Core
Designing and building large-scale training infrastructure for multi-modal generative models (video, image, audio) to support AIGC products.
Role type
Senior IC Machine Learning Infrastructure Lead (Large-scale Training)
Builds
Ultra-large scale training infrastructure for video, image, and audio generation models
Domain
Generative AI, Multi-modal AIGC, Distributed Systems
Deliverable
production ML models
Required skills
Distributed training architecture, GPU cluster management, High-performance computing, System scalability, C++/Linux development, PyTorch, Python, Multi-modal model training, Cluster resource scheduling, Performance optimization
Preferred skills
Team building, Technical strategy, Cross-functional collaboration
Technologies
PyTorch, C++, Linux, GPU clusters
Responsibilities
Designing overall architecture for foundation model training systems, Developing scalable ML systems for thousand+ GPU clusters, Researching and implementing frontier training technologies, Collaborating with algorithm teams on joint optimization, Managing distributed training and data management modules, Building and mentoring technical teams
Seniority
Senior, hands-on IC with management responsibilities
