多模态算法工程师（J93378）
Core
Research and develop large-scale pre-trained models for text, image, and video; build intelligent agents for commercial marketing video generation; optimize long-video quality and consistency.
Role type
Senior multimodal algorithm engineer (video generation & large models)
Builds
Intelligent video creation agents, high-quality video scripts, and multimodal content generation/retrieval systems
Domain
AI/ML, Computer Vision, Video Generation, Large Language Models
Deliverable
production ML models
Required skills
Python, PyTorch, TensorFlow, PaddlePaddle, pre-trained model architecture, large-scale model pre-training, diffusion models, Hadoop, Spark
Preferred skills
Top-tier conference publications, ACM/NOI/NOIP competition awards, high leaderboard rankings, open-source project leadership
Technologies
PyTorch, TensorFlow, PaddlePaddle, Hadoop, Spark
Responsibilities
Optimize model performance for business needs; solve technical challenges in video consistency; apply latest algorithms to product development; research efficient tuning methods and data strategies
Seniority
Senior, hands-on IC