AI Agent Algorithm Engineer Graduate (Global E-Commerce, Affiliate) - 2027 Start (PhD)

About the Team

The Global E-Commerce Affiliate Algorithm team focuses on building a robust content-driven e-commerce supply ecosystem that leverages advanced technology to sustain continuous growth in GMV. Key initiatives involve developing the creator growth platform, intelligent product search and recommendation systems, and AI-driven content creation tools designed specifically for creators.

The team emphasizes enhancing creator product selection and business matchmaking by creating unified search and recommendation functions that serve various distribution channels, including the Affiliate Product Marketplace and Creator Marketplace. These technologies significantly improve the efficiency of matching and transactions between creators and merchants. They also focus on expanding the high-quality creator base by utilizing lead mining and growth task frameworks, enriching content diversity and supply.

Innovation is driven by AI-native workflows that provide creators with smart sourcing, scriptwriting, and content optimization capabilities aimed at boosting content production efficiency and sustainable business growth.

This opportunity welcomes graduates eager to explore ambitious ideas, face complex challenges, and unlock vast potential for advancement. Candidates must be able to commence employment by the end of the current year and should clearly indicate their availability and graduation date in their applications.

Job Responsibilities

• Develop and advance the agent runtime system managing creator and seller agents, coordinating skills, tools, and contextual elements.
• Engineer mechanisms for context and memory management to support long, multi-turn agent interactions, including structured note-taking, summarization, editing, masking, and just-in-time retrieval.
• Perform post-training and fine-tuning of large language models using techniques such as SFT, DPO, and reinforcement learning, while building data systems to convert real conversations into training and evaluation signals.
• Design and integrate tools, skills, and MCP connectors (tools as APIs) and manage extensive inventories through effective search solutions.
• Develop evaluation methods incorporating human-agreement calibration, regression, safety monitoring, cost and latency considerations to bridge offline and online performance.
• Create self-improving systems using replay mechanisms, simulations, automated root cause analyses, and policy updates to continuously enhance performance.
• Take full ownership of a crucial end-to-end platform feature and deploy it to production across over 30 languages, monitoring business metrics like GMV, content quality, and satisfaction among creators and merchants.

Minimum Qualifications

• Completion or near-completion of a PhD in Computer Science, Artificial Intelligence, Mathematics, or related quantitative fields.
• Proficiency in Python and experience with at least one additional language among C++, Go, Rust, or Java.
• Strong foundation in machine learning, deep learning, and natural language processing, supported by hands-on engagement with large language models through academic coursework, research, internships, competitive programming, open-source work, or significant personal projects.
• Capability to interpret academic papers or engineering blogs and translate concepts into functional code.

Preferred Qualifications

• Practical experience building runtime environments, not limited to API usage, including custom agent loops, memory and context management, retrieval augmented generation, or fine-tuned models—even at research or hobby scale.
• Hands-on expertise in areas such as advanced post-training methods (SFT, DPO, RLHF, etc.), agent architectures, evaluation frameworks using LLMs as judges, inference optimizations (vLLM, TensorRT), or multilingual NLP.
• Recognized achievements through publications, high rankings in programming or machine learning competitions, or significant open-source contributions.

About TikTok

TikTok is the premier platform for short-form mobile videos, aiming to inspire creativity and deliver joy worldwide. With headquarters in Los Angeles and Singapore and multiple offices globally, TikTok fosters authentic self-expression and community connection through innovative products.

Why Join Us

We center our work on motivating creativity and impact, embracing curiosity and humility. Our strong, diverse teams learn and innovate collectively, maintaining a persistent "Always Day 1" culture that drives meaningful breakthroughs for users and the company alike. Together, we enable limitless possibilities.

Diversity & Inclusion

TikTok prioritizes an inclusive workplace where diverse skills, backgrounds, and perspectives are celebrated and nurtured. Our goal is to reflect the vast communities we serve, fostering a passionate commitment to diversity and a harmonious work environment.