Agent技术专家-剪映CapCut（深圳/广州）
Core
Design and develop end-to-end AI Agent models and evaluation systems for video and image creation, editing, and generation within CapCut.
Role type
Senior IC machine-learning engineer (AI Agent)
Builds
AI-driven video and image generation, editing, and content creation features
Domain
Consumer media / Generative AI
Deliverable
production ML models
Required skills
Large model training and optimization, PyTorch, TensorFlow, LLM/VLM post-training (SFT, RLHF), Prompt engineering, Context engineering, DeepSpeed, Megatron
Preferred skills
End-to-end AI Agent application development from 0-1, Data-model co-evolution feedback systems
Technologies
PyTorch, TensorFlow, DeepSpeed, Megatron, LLMs, VLMs
Responsibilities
Build and iterate evaluation systems for Agent models and chains; Enhance Agent capabilities using advanced algorithms; Implement innovative technical solutions for business scenarios; Explore breakthroughs in AI for video and image creation.
