Recommendation Large Model Researcher - Global E-commerce

Role Overview

This position is centered on advanced recommendation research for the International E-commerce Mall. The team works on information-feed recommendations across important user journeys, including the mall homepage, shopping funnels, product detail pages, and store or showcase surfaces. The broader goal is to serve hundreds of millions of users every day with highly relevant recommendations for products, live streams, and short videos while solving difficult problems in modern recommendation systems.

Project Goal

The assignment is to investigate fresh large-model approaches for recommendation and to move beyond long-established recommendation model structures and infrastructure designs. The expectation is to outperform existing baseline systems by a meaningful margin and to adapt the resulting methods for multiple product ecosystems, including short video, LIVE, e-commerce, and news scenarios.

Key Technical Challenges

Building large models for recommendation brings a distinct set of issues, including high engineering-efficiency demands, deeply personalized user experiences, and the need to represent media-rich content such as short videos and live streams effectively. The research effort will focus on advancing model scaling, user and content representation, multimodal understanding, long-sequence modeling, and generative recommendation methods.

Research Areas

• Content-understanding and user-behavior-driven representation learning
• Scaling recommendation model parameters and compute
• Modeling ultra-long user interaction sequences
• Generative approaches for recommendation

Responsibilities

• Conduct research on large-model methods for recommendation systems in e-commerce and related business scenarios.
• Explore new model architectures and infrastructure ideas that improve on current baseline performance.
• Study and develop techniques for content understanding, user behavior modeling, and multimodal representation.
• Investigate scalable training and inference strategies for large recommendation models.
• Design approaches for ultra-long sequence learning and generative recommendation.
• Translate research findings into solutions that can be applied across multiple product surfaces and business lines.

Requirements

• Preferably a PhD in artificial intelligence, computer science, mathematics, or a closely related field.
• Strong grounding in machine learning, along with solid programming ability.
• Hands-on research background in machine learning, NLP, computer vision, or similar areas.
• Good command of major data structures and algorithms.
• Prior involvement in search, advertising, recommendation, or large-model projects is a plus.
• Publications at top conferences such as KDD, SIGIR, RecSys, ACL, or NeurIPS are preferred.
• Strong analytical thinking, problem-solving ability, and enthusiasm for tackling difficult technical challenges.

About the Company

TikTok is a global short-form video platform with a mission to inspire creativity and bring joy. Its global headquarters are in Los Angeles and Singapore, and it also maintains offices in New York City, London, Dublin, Paris, Berlin, Dubai, Jakarta, Seoul, and Tokyo.

Why Join

The organization emphasizes curiosity, humility, resilience, and an “Always Day 1” mindset. It aims to create meaningful impact through rapid innovation, teamwork, and continuous iteration. The workplace also values inclusive collaboration and the diverse perspectives of employees from around the world.

Diversity and Inclusion

The company is committed to building an inclusive environment where people are valued for their skills, experience, and unique viewpoints. It aims to reflect the diverse communities it serves and welcomes applicants who share that commitment.