AI Agent风控算法工程师（国际化电商） - 风控
Core
Build and deploy LLM/MLLM and Agentic AI systems for risk detection and governance in e-commerce scenarios, serving 1 billion+ users across content, transaction, and traffic contexts.
Role type
Senior IC AI Agent Risk Control Algorithm Engineer
Builds
Next-generation intelligent risk control infrastructure, domain-specific risk agents, and fused traditional/LLM risk models for e-commerce
Domain
E-commerce, Anti-fraud, AI/ML
Deliverable
production ML models
Required skills
LLM/MLLM training and fine-tuning, SFT, RL, RAG, long-context handling, Agent system development, Agent evaluation (Harness Engineering), Graph models, Sequence modeling, Multi-task learning, Python/Go/C++/Java, PyTorch
Preferred skills
E-commerce or anti-cheat experience, Black/gray market countermeasure knowledge, English communication (IELTS/TOEFL 6+)
Technologies
Qwen, DeepSeek, Gemini, PyTorch, CPT, SFT, RL, RAG
Responsibilities
Develop risk detection models for marketing, transactions, merchants, and malicious buyers; Optimize risk control domain large models using CPT/SFT/RL; Build Agentic systems for automated risk analysis and strategy application; Fuse large models with traditional models to improve accuracy under cost constraints; Construct domain datasets and evaluation benchmarks for continuous iteration.
