I am Maolei Huang, an algorithm expert and researcher at Alibaba in Beijing. My work focuses on large-scale ranking and recommendation systems, supply and revenue optimization, and LLM-powered ranking and agent systems, covering industrial AI from problem formulation, feature and model design, and training optimization to online serving, experimentation, and continuous iteration under strict latency and reliability constraints.

My research interests include generative ranking, user behavior modeling, temporal and delayed-feedback learning, LLM alignment, reinforcement learning, multi-agent collaboration, and multi-objective optimization. I received my M.S. from UCAS/ICT, CAS and my B.S. from Shandong University. My work has been published at KDD, AAAI, SIGIR, ICLR, IEEE IoT-J, and ICDM, including first-author publications at KDD, AAAI, and SIGIR, along with granted invention patents. I also serve as a reviewer for KDD, RecSys, and SIGIR. My goal is to turn real-world challenges into algorithmic innovations and translate advances in ranking, recommendation, LLMs, and agents into scalable industrial systems.

我是 Maolei Huang，目前在阿里巴巴北京担任算法专家与研究员，主要从事大规模排序与推荐系统、供给与收入优化，以及基于大模型的排序与智能体系统。我的工作覆盖工业级 AI 的端到端落地，包括问题建模、特征与模型设计、训练优化、在线服务、实验评估和持续迭代，并长期面对低延迟与高可靠性约束。

我的研究兴趣包括生成式排序、用户行为建模、时序与延迟反馈学习、LLM 对齐、强化学习、多智能体协作和多目标优化。我硕士毕业于中国科学院大学/中科院计算所，本科毕业于山东大学。相关研究成果发表于 KDD、AAAI、SIGIR、ICLR、IEEE IoT-J 和 ICDM，包括 KDD、AAAI 和 SIGIR 的第一作者论文，并获得多项发明专利授权；同时担任 KDD、RecSys 和 SIGIR 审稿人。我的目标是将真实业务挑战转化为算法创新，并把排序、推荐、大模型和智能体方向的研究进展落地为可扩展的工业系统。