高级广告审核Agent算法工程师-深圳/北京
Core
Design and implement core frameworks for review Agents with capabilities in rule understanding, case reasoning, autonomous reflection, and continuous iteration to identify complex risk content and provide explainable judgments.
Role type
Senior IC Agentic AI Engineer (Content Safety/Ads)
Builds
Agentic review systems, multi-agent collaboration pipelines, and automated content repair workflows
Domain
Internet / Content Safety / Ad Tech
Deliverable
production ML models
Required skills
Large model training and fine-tuning, Reinforcement Learning (GRPO/OPD), Harness Engineering, Multi-Agent systems, Planning, Reflection, Memory, SFT, NLP, Machine Learning
Preferred skills
Content safety or ad algorithm business experience, Top-tier conference publications (NeurIPS/ICML/ICLR/CVPR/KDD/ACL), Independent Agent system development
Technologies
SFT, GRPO, OPD, AIGC, Multi-modal generation
Responsibilities
Design and implement core review Agent frameworks; Deploy Agentic technologies for content safety; Build recursive self-iteration loops for model improvement; Generate content repair outputs using Agent planning and multi-modal generation.
