地图商业化算法工程师（J105420）
Core
Develop and optimize commercialization algorithms for search and recommendation scenarios to enhance ad revenue, user experience, and ecosystem health.
Role type
Senior IC machine-learning engineer (advertising & recommendation)
Builds
Ad ranking systems, bidding mechanisms, and multi-objective estimation models for search and recommendation platforms.
Domain
Internet advertising, search, and recommendation systems.
Deliverable
production ML models
Required skills
CTR/CVR estimation, ranking learning (LTR), multi-task/multi-objective modeling, deep learning, C++/Go, TensorFlow/PyTorch, large-scale sparse feature modeling, large model SFT fine-tuning, big data processing (Spark/Hadoop/Flink), A/B testing design, causal inference.
Preferred skills
Reinforcement learning, PID control systems, sequence modeling, business acumen.
Technologies
C++, Go, TensorFlow, PyTorch, Spark, Hadoop, Flink.
Responsibilities
Design and optimize ad bidding mechanisms and pricing strategies (oCPC/oCPM); lead iteration of deep learning models for CTR/CVR prediction; balance user experience with commercial value through ad-natural result fusion and frequency control; execute end-to-end optimization from data to online service.
Seniority
Senior, hands-on IC
