阿里巴巴Research Intern-Post training
任职要求
● PhD, Master's, or Bachelor's degree in Computer Science, Engineering, Mathematics, or related fields.
● Ability to implement and verify state-of-the-art NLP or Computer Vision algorithms.
● Experience in successfully applying machine le…工作职责
【About our team】 Our team is dedicated to revolutionizing B2B e-commerce through the development of a groundbreaking AI search product. Leveraging cutting-edge LLM technology, we're creating a brand new AI search engine and streamline the entire purchasing process for B2B customers. 【Responsibilities】 ● Deliver full projects by defining the data structure, framework, design, and evaluation metrics for AI solution development and implementation. ● Develop and deploy Large Language Models (LLMs) to empower search and agent systems for a variety of applications. ● Identify new and upcoming product innovation areas by interacting with potential external and internal collaborators. ● Identify undefined problems in existing technology and engage stakeholders and leaders to address them.
1. 研究和打造新一代具身智能体（AI Agent），推动AI从“理解世界”迈向“自主思考和持续学习”。 2. 为人形机器人、虚拟人、语音助手等场景设计具备长期记忆、推理、规划、工具调用和环境互动能力的智能系统。 3. 构建端到端的智能体学习框架
我们致力于构建下一代 空间智能（Spatial Intelligence）系统，让AI不仅能“看懂世界”，更能理解空间结构、推理物体关系、规划行动轨迹，并在虚拟或真实环境中持续学习与演化。 你将与团队一起： 研发具备空间理解、物体感知、轨迹预测与交互规划能力的智能体模型； 构建融合 视觉语言模型（VLM）与世界模型（World Model） 的系统，实现3D场景、深度、物理与可供性（Affordance）的联合建模； 使用 Game Engine（Unreal / Unity / Isaac Sim） 搭建高保真虚拟环境，用于数据生成与智能体评测； 基于 vLLM / Ray 构建高效多模态数据管线，实现大规模生成、自动标注与验证； 推动空间智能在机器人与具身智能领域的应用落地。
① 协助完成跨境电商相关业务的B端用户研究工作，挖掘产品体验痛点及优化机会点； ② 协助完成跨境电商相关业务的行业市场/趋势研究、C端消费者研究，赋能产品规划方向； ③ 研究内容包括但不限于：行业研究、问卷调研、商家访谈、用户画像/体验研究、产品数据追踪等，在指导下输出研究报告。
As a research intern on the embodied AI projects, you will help build the advanced robotic policies with humanoid robots for the challenging industrial scene — enabling sim-first development, real-world deployment, and continuous learning to make them smarter over time. The ideal candidate will have strong research skills for robotics. What You Will Be Doing: • Bring the latest advancements in embodied AI to simulated and real humanoid robots, successfully apply research in the challenging industrial scenes, like healthcare scenarios. • Build sim-first robot policies with SimReady digital twin and minimize the sim2real gap. • Collaborate across team boundaries to integrate NVIDIA robotics products such as Jetson Thor, Isaac GR00T, Cosmos, and Isaac Sim/Lab into the research work. • Take on a variety of challenges, solve practical research topics. • Deploy and test the developed software on real humanoid robots.