Large Recommendation Model Algorithm Engineer - Global E-Commerce

About the Team

The E-commerce Recommendation Foundation team focuses on developing advanced recommendation intelligence by creating a unified Foundation Model to power diverse business scenarios and recommendation tasks, from retrieval to ranking and re-ranking. Their work aims to revolutionize recommendation systems through a generative paradigm driven by event sequences, incorporating large language models (LLMs), multimodal understanding, reinforcement learning, and system optimization.

The team encourages innovative research and engineering excellence, fostering an open environment where every member can hypothesize and validate ideas to shape the future of recommendation systems, targeting general intelligence capabilities.

Responsibilities

• Develop and enhance cross-scenario shared Foundation Models for unified modeling and efficient inference, advancing the event-sequence-driven generative recommendation paradigm with multimodal and generative features.
• Utilize LLM technologies across retrieval, ranking, and re-ranking phases, including involvement in model training, inference optimization, and collaborative system design.
• Investigate ways to integrate LLMs and visual language models (VLMs) with recommendation systems to create adaptable and continually improving intelligent recommenders.
• Research and develop end-to-end generative recommendation approaches and optimize systems balancing performance and user experience.

Qualifications

Minimum Requirements:

• Strong theoretical knowledge of machine learning, deep learning, or information retrieval.
• Proficiency in Python programming and familiarity with popular deep learning frameworks such as PyTorch.
• Keen interest in intelligent recommendation systems and a self-motivated research attitude.

Preferred Qualifications:

• Experience working on large-scale recommendation systems or training large models, with significant technical contributions in specific areas.
• Research or published work in large language models, multimodal learning, reinforcement learning, or generative recommendations.
• Understanding of pre-training and fine-tuning procedures for large language or foundation models.

About ByteDance

Established in 2012, ByteDance's mission is to inspire creativity and enrich life through an extensive suite of global products including TikTok, Lemon8, CapCut, Pico, and China-focused platforms like Toutiao and Douyin. The company strives to make content creation, consumption, and connection more engaging and accessible worldwide.

Why Join ByteDance

ByteDance aims to foster creativity by building innovative products that empower authentic self-expression and global connection. The company values curiosity, humility, and impact, promoting a culture of continuous innovation with an "Always Day 1" mindset to drive meaningful breakthroughs. By collaborating within diverse and inclusive teams, ByteDance offers limitless opportunities for professional growth and success.

Diversity & Inclusion

ByteDance is dedicated to fostering an inclusive workplace where employees’ diverse skills, experiences, and perspectives are respected and celebrated. Reflecting its global platform, the company prioritizes diversity and aims to create an environment that supports and uplifts all community voices.