Research Engineer (AI Infrastructure) - Large Model Applications - Global Frontier Tech Recruitment Program - 2027 Start (PhD)

Overview

We are inviting driven individuals to join our team in 2027 to innovate boldly, tackle complex engineering challenges, and accelerate career growth in an inspiring environment. Candidates must be available to onboard by the end of 2027 and clearly state graduation and availability dates on their resume.

Team Role

This role is part of our team developing TikTok's Recommendation and Search system architecture. We focus on maintaining system stability and high availability, optimizing online and offline data streams, resolving bottlenecks, and reducing operating costs. The team also constructs core system components such as recommendation and data middle offices to support rapid product incubation and business-to-business services.

Technical Context

Our systems face increasing complexity in search, advertising, and recommendation applications. Large models enhance personalization and content relevance but require real-time responsiveness, reliability, and scalability. Challenges include distributed training, inference acceleration, heterogeneous hardware utilization, and multimodal data processing.

Project Focus

This area addresses overcoming critical infrastructure bottlenecks caused by the expansion of model scales and multimodal data volume. The goal is to engineer efficient, reliable, and scalable infrastructure solutions to support large models, which underpin search, advertising, and recommendation functionalities.

Technical Challenges

• Redesigning native training and inference architectures for large language models (LLMs)
• Extreme performance tuning and innovation in AI infrastructure
• Innovating end-to-end generative AI paradigms
• Developing multimodal audio-visual data infrastructure and enhancing data quality
• Advancing multimodal data representation and retrieval-augmented generation (RAG) based systems

Impact

Building cutting-edge generative AI infrastructure will enable TikTok’s search, advertising, and recommendation services to handle ultra-long context sizes, achieve millisecond-level response times, and highly accurate information understanding, propelling intelligent business upgrades.

Candidate Qualifications

• PhD completed or soon to be completed in Artificial Intelligence, Computer Science, Computer Engineering, or related fields
• Strong programming skills and mastery of data structures and fundamental algorithms
• Proficiency in C/C++ for traditional coding tasks or Python for intelligent coding tasks, including implementing complex algorithms and model iterations
• Engineering mindset balancing system performance and cost effectiveness

Preferred Qualifications

• Publications in top-tier (CCF-A) journals or conferences such as VLDB, SIGIR, or OSDI are advantageous
• Resilience, excellent communication, teamwork skills, passion for technology, and eagerness to innovate and face challenges collaboratively

About TikTok

TikTok is a global leader in short-form mobile video, dedicated to inspiring creativity and bringing joy. Our worldwide offices encompass Los Angeles, Singapore, New York City, London, Dublin, Paris, Berlin, Dubai, Jakarta, Seoul, and Tokyo.

Why Join Us

We foster a culture of curiosity, humility, and impact-oriented collaboration within a rapidly expanding tech company. We embrace challenges as growth opportunities and adopt an iterative, "Always Day 1" mindset, achieving significant technical breakthroughs together.

Diversity & Inclusion

We are committed to cultivating an inclusive workplace that values diverse skills, experiences, and perspectives. Our mission connects global users and reflects a rich variety of communities, celebrating all voices.