Xiaosen Chen （陈潇森）

Algorithm Engineer, T4 · Recommendation Systems

算法工程师 T4 · 推荐系统

Video Feed Strategy & Ranking · 2024.07 – Present

视频 Feed 策略与排序 · 2024.07 – 至今

• Drove core video distribution metrics through iterative strategy optimization, owning experiments end-to-end from design to launch.
• 主导视频分发核心指标策略迭代，负责实验全链路设计与上线。
• Led transition from rule-based strategies to model-centric ranking, improving distribution efficiency at scale.
• 推动从规则驱动向模型中心化排序转型，规模化提升分发效率。
• Applied LLM techniques (SFT/DPO) to recommendation workflows for content understanding and user intent modeling.
• 将大模型技术（SFT/DPO）引入推荐链路，探索内容理解与用户意图建模落地。

Recommendation Infrastructure · 2023.07 – 2024.06

推荐基础设施 · 2023.07 – 2024.06

• Built and maintained core recommendation pipeline components; contributed to feature engineering and recall optimization.
• 负责推荐核心链路组件开发与维护，参与特征工程与召回优化。
• Developed tooling to accelerate offline evaluation and A/B experiment analysis.
• 开发离线评估与 A/B 实验分析工具，提升团队实验迭代效率。