About the Team We are the User Growth Algorithm Team for Global E-Commerce — the core engine driving the explosive growth of TikTok's global e-commerce business. Here, we go beyond solving conventional recommendation and user reach problems; we are committed to reshaping the shopping experience of global consumers through cutting-edge promotion algorithms. As the Promotion Algorithm Team, we command massive budgets and enormous traffic flows. Through state-of-the-art algorithmic modeling and operations research optimization, we strive to achieve globally optimal ROI in budget allocation—directly driving scaled growth in overall MAC (Monthly Active Customers) and GMV. Here, every line of code you create will influence the purchasing decisions of tens of millions of users worldwide, and every model iteration will unlock tremendous business value. By joining us, you will have the opportunity to build on the unique ecosystem that fuses content with e-commerce to solve rare industry challenges in the joint optimization of "content recommendation + real-time subsidies."
We are looking for talented individuals to join us for an internship. PhD internships at Our Company provide students with the opportunity to actively contribute to our products and research, as well as to the organization's future plans and emerging technologies. Our dynamic internship experience blends hands-on learning, enriching community-building and professional development events, and collaboration with industry experts. Applications will be reviewed on a rolling basis, so we encourage you to apply early. Please clearly state your availability in your resume (Start date, End date).
Responsibilities
- Precision Subsidies & Intelligent Pricing: Model user demand elasticity across products and coupon values, combined with operations research optimization to solve for optimal budget allocation under strict constraints — maximizing GMV and conversion rate.
- Intelligent Global Budget Optimization (Operations Research & Game Theory): Design globally optimal allocation and pricing algorithms based on elasticity modeling, enabling large-scale budgets to flow precisely across billions of users, massive product catalogs, and diverse scenarios for optimal budget efficiency.
- Deep Synergy Between Traffic and Marketing (Vouchers × Traffic): Integrate real-time item-wise voucher capabilities into recommendation pipelines across short videos, live streaming, and marketplace. Break down barriers between marketing and recommendation to enable joint "traffic + subsidy" decision-making and global optimization.
- Next-Gen Marketing Tech: Go beyond traditional discriminative uplift prediction. Research and productionize cutting-edge causal inference and AI technologies — including Uplift Modeling, E2E Modeling, RL, LLMs, and AI Agents — to directly address user conversion and repeat purchase behavior.
Minimum Qualifications
- Currently Pursuing a PhD in Software Development, Computer Science, Computer Engineering, or a related technical discipline.
- Strong expertise in machine learning and deep learning, with familiarity in commonly used models for recommendation, advertising, search, or marketing.
- Hands-on experience and deep understanding in at least one of the following areas:
- Causal Inference / Uplift Modeling: Familiarity with various uplift modeling approaches; candidates with industrial deployment experience are preferred.
- Operations Research Optimization: Familiarity with linear programming, integer programming, dynamic programming, MCKP, etc.; experience in budget allocation or pricing algorithms is preferred.
- Search, Recommendation & Advertising Algorithms: Experience with large-scale CTR/CVR estimation, multi-objective optimization, or real-world reinforcement learning applications is preferred.
- Frontier AI Technologies: Experience applying LLMs, AI Agents, or Generative Recommendation in business scenarios is preferred.
- Excellent business acumen and logical thinking, with the ability to translate complex business problems into mathematical or algorithmic models and drive them into production.
Preferred Qualification
- Publications in top-tier venues such as KDD, NeurIPS, ICML, IJCAI, or AAAI, or outstanding results in data mining competitions such as Kaggle, will be a strong plus.