大模型算法工程师（图片美化和创作方向）-剪映CapCut
Core
Develop algorithms and applications for image beautification and creation agents, focusing on data mining, synthesis, and strategy optimization for large-scale image creators.
Role type
Senior IC large model algorithm engineer (image generation & editing)
Builds
CapCut image product features and creation agents
Domain
Consumer media / Large language and vision models
Deliverable
production ML models
Required skills
Large model training (SFT, RLHF), multimodal perception, computer vision, NLP, data synthesis, agent optimization
Preferred skills
Top-tier competition awards (ACM/ICPC, Kaggle), publications in top conferences (ACL, CVPR, ECCV), prior experience in image/video generation models
Technologies
Transformer, DiT, ViT, Self-Instruct
Responsibilities
Research high-quality image creation data mining and synthesis; Explore post-training algorithms for full-modal large models; Optimize agent end-to-end workflows; Deploy technical solutions to the CapCut image product.
Seniority
Senior, hands-on IC
