Machine Learning Engineer (Recommendation) - Global E-Commerce
Location
singapore, singapore
Job Type
Full-time
Category
Other-General
Posted
June 11, 2026
Responsibilities
About the team Our team works on large-scale recommendation systems for various offerings under ByteDance and its affiliates, focusing on developing recommendation algorithms/models/strategies. We are committed to developing cutting-edge solutions for e-commerce recommendation systems. Responsibilities - Work on recommendation systems, involving contents of various forms ranging from products, short videos to live streams, with each unified recommendation model fulfilling heterogeneous E-commerce scenarios/goals across multiple countries. - Optimize e-commerce recommendation models at massive scales, using deep learning/transfer learning/multi-task learning techniques. - Data mining and analysis to improve the quality of recommended contents. - Conduct research on various topics, which aim to optimize content recommendation circulation, ranging from ensuring diversity and new discovery in recommendation contents, to cold-start problem for new users/items and discove...
About the team Our team works on large-scale recommendation systems for various offerings under ByteDance and its affiliates, focusing on developing recommendation algorithms/models/strategies. We are committed to developing cutting-edge solutions for e-commerce recommendation systems. Responsibilities - Work on recommendation systems, involving contents of various forms ranging from products, short videos to live streams, with each unified recommendation model fulfilling heterogeneous E-commerce scenarios/goals across multiple countries. - Optimize e-commerce recommendation models at massive scales, using deep learning/transfer learning/multi-task learning techniques. - Data mining and analysis to improve the quality of recommended contents. - Conduct research on various topics, which aim to optimize content recommendation circulation, ranging from ensuring diversity and new discovery in recommendation contents, to cold-start problem for new users/items and discove...