生成式排序模型算法工程师（J104592）
Core
Develop generative ranking models for recommendation and advertising scenarios, applying Transformer and sequential modeling to CTR/CVR/GMV estimation.
Role type
Research-level algorithm engineer (generative AI/recommendation systems)
Builds
Generative ranking models, multimodal large model experiments, and optimized inference pipelines for ad platforms
Domain
Internet advertising, recommendation systems, generative AI
Deliverable
production ML models
Required skills
Transformer architectures, sequential modeling, Python, PyTorch or TensorFlow, deep learning fundamentals, data structures
Preferred skills
Large-scale distributed training, long-sequence modeling, semantic ID research, reinforcement learning, top-tier conference publications
Technologies
Transformer, DeepFM, DIN, GPT, LLM, PyTorch, TensorFlow
Responsibilities
Research and develop generative ranking models for ad scenarios, conduct experiments on multimodal large models, optimize model training and inference, reproduce and evaluate state-of-the-art algorithms, collaborate on data processing and A/B testing
Seniority
Junior (2027 graduate), research-focused IC
