Senior Machine Learning Engineer (CV/NLP/Multimodal/LLM) -E-commerce Governance

About the Team The Governance and Experience Algorithm team was established in September 2020 and its main job is to support the following businesses with the most advanced AI technology： - Combat any kinds of risks/violations issues in E-commerce scenarios. - Build an E-commerce ecosystem and improve platform service capabilities

Responsibilities: - Responsible for identifying algorithms for risk/violation/low-quality issues in e-commerce scenarios such as products, shopping cart short videos, and live streaming with products. - Responsible for mining and identifying algorithms to detect fraudulent behaviors and identify risky, violated, or low-quality merchants and creators. - Responsible for optimizing algorithms for large-scale brand libraries, products, and content deduplication/clustering. - Responsible for data construction, instruction fine-tuning, CoT, alignment, and other work for large models in the e-commerce domain, aiming for ultimate effect optimization in the e-commerce domain. - Responsible for exploring reinforcement learning and operational research algorithms in intelligent dispatching and intelligent audit scheduling to improve audit quality and efficiency. What You Will Do: - Optimize various basic algorithms such as NLP, vision, multimodal, search, graph, LLM, etc. to provide support for governance business, explore cutting-edge technologies, and apply them to e-commerce business scenarios. - Streamline the optimization and iteration process for computer vision models in e-commerce settings, including fine-grained classification of product images, product object recognition, product subject recognition, feature extraction, logo detection, brand recognition, etc. - Optimize video classification, multimodal content mining, and multimodal content understanding of e-commerce short videos and live broadcasts, and optimize the e-commerce short video shopping experience. - Optimize the risk identification model related to text content in e-commerce scenarios, such as product copywriting, picture text, and voice descriptions. - Optimize the risk mining algorithm for comments, reports, and other public opinion content related to products/short videos/live broadcasts, and improve the platform governance effect. - Responsible for the research of large-scale multimodal (vision, speech, natural language, etc.) algorithms and their implementation in e-commerce scenarios. - Extract extensive data from various entities, such as content in e-commerce live broadcasts, products, merchants, and influencers. Model large-scale networks to support business scenarios like content understanding, multimodal representation, and community mining for problem-solving purposes. - Participate in the construction of a large-scale graph storage and graph learning platform, supporting the basic ability of graph learning with billions of nodes and edges.

Minimum Qualifications: - Bachelor's degree or above in computer science or related field - Solid coding skills, ability to develop in a Linux environment, proficient in Python, Go, or C++ - Solid foundation in data structures/algorithms, proficient in machine learning/deep learning theory, and rich practical experience - Familiar with 1-2 areas in natural language processing, computer vision, multimodal, graph algorithms, search algorithms, text/data mining, and LLM - Excellent analytical and problem-solving skills, passionate about challenging problems - Good team spirit and strong communication skills

Preferred Qualifications - Experience in platform governance/audit/risk control/content security or overseas internet/e-commerce business is preferred - Experience studying, working, or living overseas is preferred.