13 days ago
Beijing, ChinaStaff+
Responsibilities
- Diagnose bottlenecks and quantify defects across the search and recommendation funnel.
- Architect and implement end-to-end machine learning, deep learning, generative AI, and LLM solutions for search.
- Build query understanding services for intent detection, query rewriting, semantic expansion, and entity extraction.
- Design and scale semantic and dense retrieval systems using traditional vector search and generative retrieval.
- Develop LLM-based listwise and pairwise reranking models and optimize real-time serving trade-offs.
- Optimize offline and online inference using model distillation, quantization, vLLM, and TensorRT.
- Conduct rigorous evaluations and A/B tests for production search and recommendation systems.
- Mentor senior engineers and drive the search AI roadmap across Infrastructure, Platform, Product, and other organizations.
Requirements
- 10+ years of professional experience in Search, Information Retrieval, recommender systems, machine learning, deep learning, or natural language processing.
- Proficient software engineering and coding skills in Python, C++, or Java.
- Experience handling large data volumes, training large-scale models, and serving high-QPS pipelines in distributed environments such as Spark, Ray, or Kubernetes.
- Ability to formulate ambiguous business problems mathematically and deliver production-grade machine learning solutions.
- Master’s degree in Computer Science, Data Science, or a related engineering field.
- Preferred: hands-on experience fine-tuning, aligning, or distilling LLMs for search, recommendation, or NLP tasks, including SFT, LoRA, RLHF, DPO, query expansion, dense retrieval, or reranking.
Benefits
- The recruitment process includes application review, job fit interview, focus interview, and offer, with scheduling subject to change.
- The posting may close early if all openings are filled.
- Eligible veterans and people with disabilities may receive preferential treatment in accordance with applicable laws.
Tech Stack
Categories
About Coupang
Coupang builds and operates a South Korea–focused e-commerce marketplace with an end-to-end logistics network (Rocket Delivery), plus food delivery, video streaming, and fintech under brands such as Coupang, Eats, and Play. Revenue comes from first-party retail, third-party marketplace services, advertising, and memberships (Rocket WOW). Founded in 2010, the company is headquartered in Seattle and is publicly listed on the NYSE (CPNG).