内容理解算法工程师（J100150）
Core
Develop and research multimodal content understanding algorithms for search and recommendation scenarios, covering semantics, quality, and interest dimensions across images, video, and web content.
Role type
Senior IC multimodal content understanding algorithm engineer
Builds
Multimodal content understanding systems integrated with search and recommendation pipelines
Domain
Search and recommendation, multimodal AI
Deliverable
production ML models
Required skills
Image/video understanding, object detection, segmentation, OCR, semantic understanding, multimodal large models, C/C++, Python, PyTorch, PaddlePaddle
Preferred skills
Experience with multimodal large models
Technologies
PyTorch, PaddlePaddle
Responsibilities
Research and develop multimodal content understanding algorithms; Optimize recall, ranking, and fusion stages by integrating content understanding with search/recommendation; Explore and apply LLMs and generative AI to content understanding; Conduct technology research and algorithm innovation in content understanding
Seniority
Senior, hands-on IC
