Machine Learning Engineer - Logistics Fulfillment

About our Team We are the supply chain and logistics algorithm team under Global E-commerce. We focus on technical optimization of the global logistics and fulfillment process. Through the in-depth integration and innovation of cutting-edge AI algorithms such as reinforcement learning and large models, combined with operations research optimization technologies, we drive logistics cost reduction, service efficiency improvement, and user experience enhancement. Our team is currently expanding rapidly with locations in Hangzhou, Shanghai, Singapore, and Seattle. We focus on the algorithm implementation and breakthroughs of the self-owned logistics system, deeply participate in the full-link intelligent optimization projects from warehouse distribution to last-mile delivery, and focus on tackling the implementation of core directions such as reinforcement learning in dynamic and real-time logistics scenario decision-making, and large models in complex supply chain deduction and intelligent interaction. We continue to empower the global e-commerce logistics business with cutting-edge algorithms. Job Responsibilities 1. Responsible for the implementation of operations research and AI projects related to self-owned warehouses and quasi-self-owned warehouses, promote the algorithm research and development of in-depth cooperative innovation warehouses with service providers. Apply deep reinforcement learning to core in-warehouse links such as intelligent order batching, wave picking, and dynamic packaging optimization, complete the implementation and iteration of internal systems, and break through the efficiency bottleneck of traditional operations research optimization in complex dynamic scenarios. 2. Take charge of the algorithm construction for First-Mile/Last-Mile scenarios. Based on the reinforcement learning framework, make decisions on dynamic collection route planning, intelligent merchant onboard/offboard, and real-time regional division. Iterate the platform's intelligent order allocation algorithm and the GTL three-segment code prediction model at the delivery station level. Implement a LLM-driven intelligent labor scheduling optimization plan, which automatically generates scheduling strategies combining business rules and real-time capacity fluctuations to improve the efficiency of last-mile fulfillment. 3. Participate in global logistics backbone network planning projects, build the ability to simulate complex supply chain scenarios based on large language models, accurately assess the changes in timeliness, traffic and platform profits corresponding to changes in warehousing and distribution, deepen the implementation of large models in multi-scenario simulation and automatic parameter optimization in network planning scenarios, and optimize the decision-making efficiency and scientific nature of global logistics network layout. 4. Responsible for the research and development of intelligent scheduling algorithms in Middle-Mile scenarios. Lead the algorithm design and implementation of cross-regional logistics projects such as the US TPE(Transportation Planning and Execution) project. Apply reinforcement learning to dynamic route planning and capacity scheduling for trunk line transportation, and continuously optimize the cost and timeliness stability of trunk transportation. 5. Explore the implementation of AI innovation projects. In addition to building the ability to recognize images scanned in the warehouse, we will focus on promoting the application of multimodal large models in intelligent identification of missing parts upon receipt and automatic traceability of abnormal scenarios. This will solve the business pain points of low efficiency of manual sampling and many disputes over compensation for abnormalities, and reduce operational risks and costs. 6. Deeply explore the full-link data of logistics, extract the core directions for optimizing logistics operation models and improving service quality. Participate in product plan design, and collaborate with cross-functional teams such as product, operation, and business analysis to promote the productization of algorithms, continuously empowering the business. Minimum Qualifications 1. Bachelor's degree or above, majoring in mathematics, statistics, computer science, industrial engineering, supply chain and logistics, operations research, or other related fields. 2. Have a solid foundation in data structure and algorithm design, and be proficient in at least one programming language such as Python, Java, C++, or R. 3. Proficient in deep learning, reinforcement learning (especially in sequence decision-making and multi-agent reinforcement learning), operations research optimization (integer programming, stochastic programming), and large models (LLM application fine-tuning, multimodal large models), with relevant algorithm implementation experience in the logistics or supply chain field preferred. 4. Have strong learning ability and curiosity, clear logic, good teamwork and communication skills, and proficient English writing and communication skills. Preferred Qualification 1. Priority will be given to candidates who have published papers or achieved project results in top computer science conferences (such as NeurIPS, ICML, ICLR, KDD, AAAI, IJCAI) or top operations research journals (such as OR, MS, MP, M&SOM, POMS, TS, TRB).