岗位职责
- 多语言和多模态大模型后训练：负责多语言基座和翻译技术的落地，实现文本、图片、语音、视频等多模态的跨语言理解，让不同语言的用户无障碍交流；
- AI Agent in 推荐算法：负责AI Agent技术在 rednote 的落地，包括用户memory、生成式召回、基于Agent的效果提优新范式等；
- AI Agent in 搜索算法：负责AI Agent技术在 rednote 的落地，方向包括 AI 搜索、生成式召回、基于 LLM 的新相关性范式等。 【About the Team】 The Global Search & Recommendation Algorithm Team is a core force behind rednote’s global expansion strategy. Our work spans key areas including community recommendation, search, AI-powered search, translation, and multilingual foundation models. In early 2025, our LLM-powered translation technology gained widespread attention, showcasing the team’s technical capabilities and impact. Global expansion calls for a fundamental rethinking of how we build products. Unlike mature businesses focused on continuous optimization, rednote offers us a genuine 0-to-1 opportunity: fewer legacy constraints, greater potential for innovation, and the chance to reshape search and recommendation algorithms through an AI-native approach. Here, you’ll tackle some of the most challenging algorithmic problems in real-world global business scenarios, including multilingual understanding, cross-cultural relevance, cold-start challenges, and cross-domain understanding. If you’re eager to explore uncharted territory and solve complex technical challenges, join us in building the next generation of community products for users around the world. 【Job Responsibilities】
- Multilingual and Multimodal LLM Post-training Drive the post-training of multilingual and multimodal foundation models, as well as the deployment of multilingual foundation models and translation technologies. Enable cross-lingual understanding across multiple modalities, including text, images, audio, and video, helping users who speak different languages communicate seamlessly.
- AI Agents for Recommendation Algorithms Drive the application of AI agent technologies to recommendation systems at rednote. Explore innovative approaches including user memory, generative retrieval, and new agent-based paradigms for improving recommendation performance.
- AI Agents for Search Algorithms Drive the application of AI agent technologies to search systems at rednote. Explore areas including AI-powered search, generative retrieval, and new LLM-based relevance paradigms.
职位要求
- 本科及以上学历，计算机、人工智能、电子信息、自动化、软件或数学等相关专业；
- 扎实的机器学习基础，在 RecSys、KDD、NeurIPS、ICML、ACL、SIGIR、ICLR 等国际顶级会议以第一作者发表过高水平论文者优先；在 KDD Cup、Kaggle、ICPC 等知名竞赛中取得领先名次者优先；
- 有大模型、搜索、推荐、NLP 或多语言相关的研究或实习经历者优先；
- 编程基本功扎实，熟悉常用数据结构与算法，擅长 Java / C++ / Python 中至少一门语言；
- 具备良好的英语沟通能力，能适应跨文化协作环境者优先；
- 踏实自驱，有较强的逻辑思维和抽象能力。 你会得到： - 完整参与一款全球化产品从 0 到 1 的算法体系搭建，方向和路径由你定义； - 多语言真实业务场景下的海量数据和前沿技术挑战，与团队共建 AI native 的搜推新范式； - 支持在顶会持续发表研究成果，与学界保持紧密连接。 【Job Requirements】
- Educational Background: Bachelor’s degree or above in Computer Science, Artificial Intelligence, Electronic Information Engineering, Automation, Software Engineering, Mathematics, or a related field.
- Strong Machine Learning Fundamentals: A solid foundation in machine learning. Candidates who have published high-quality research as first authors at leading international conferences such as RecSys, KDD, NeurIPS, ICML, ACL, SIGIR, or ICLR are preferred. Top placements in well-known competitions such as KDD Cup, Kaggle, or ICPC are also a plus.
- Relevant Research or Internship Experience: Research or internship experience in LLM, search/recommendation systems, NLP, or multilingual technologies is preferred.
- Strong Programming Skills: Solid programming fundamentals and familiarity with common data structures and algorithms. Proficiency in at least one of Java, C++, or Python.
- English Communication Skills: Good English communication skills and the ability to work effectively in cross-cultural environments are preferred.
- Ownership and Analytical Thinking: A proactive, self-driven mindset, strong logical reasoning, and the ability to think abstractly. 【What You’ll Gain】 -Build a global product from 0 to 1: Play an integral role in building the algorithmic systems of a global product from the ground up, with the opportunity to help define its technical direction and development path. -Tackle large-scale challenges and shape the future of AI-native search and recommendation: Work with large-scale datasets and cutting-edge technologies in real-world multilingual scenarios, collaborating with the team to establish new paradigms for AI-native search and recommendation. -Stay at the forefront of research: Receive support for publishing research at top-tier academic conferences and maintain close connections with the research community.
