Machine Learning Engineer Graduate (E-Commerce Content Recommendation - Generative & Large Recommendation Model) - 2027 Start (PhD)
Core
Building and rewriting an industrial recommendation system for TikTok Shop's global e-commerce video and image-text content using LLM foundations to serve hundreds of millions of users.
Role type
PhD-level Machine Learning Engineer (Generative Recommendation)
Builds
End-to-end recommendation stack (retrieval, ranking, blending) for short-video, livestream, and product scenarios
Domain
E-Commerce, Large Language Models, Recommendation Systems
Deliverable
production ML models
Required skills
LLMs/foundation models, NLP, CV, RL, recommendation/search/ads, math behind models, algorithms and data structures, problem decomposition
Preferred skills
CUDA/Triton kernel development, large-scale distributed training, LLM post-training (SFT/RLHF/DPO/GRPO), agent-system construction, inference acceleration, top-tier publications (KDD, NeurIPS, etc.)
Technologies
LLMs, MLLMs, RQ-VAE, SID, GRPO, CUDA, Triton, KV caching, speculative decoding
Responsibilities
Scale recommendation models to billions of parameters; build one-stage generative retrieval; inject world knowledge via reasoning models; optimize training/inference to hardware limits; implement coding agents for R&D; conduct original research on open problems
Seniority
PhD, Research/Engineering Lead