机器学习平台研发工程师（训推框架方向）-Data AML
Core
Design and develop machine learning engines and training frameworks for data preprocessing, training, and inference to support advertising, recommendation, and search scenarios.
Role type
Senior IC machine-learning platform engineer (training/inference frameworks)
Builds
High-performance real-time estimation systems and internal ML training platforms
Domain
Internet advertising, recommendation, and search
Deliverable
production ML models
Required skills
C++, Python, GPU programming, compiler development, distributed computing, high-performance networking, deep learning framework internals (TensorFlow/PyTorch/MXNet), operator fusion, model quantization, mixed precision training, heterogeneous hardware acceleration
Preferred skills
Real-time system design, algorithm-system joint optimization, user-facing product development experience, distributed system development
Technologies
TensorFlow, PyTorch, MXNet, CNN, RNN, LR, SVM, RF, GBDT, FM, DeepFM, DCN, xDeepFM
Responsibilities
Design and develop machine learning engines; research and develop internal ML training frameworks; design and develop real-time high-performance estimation systems; optimize performance and upgrade architecture; collaborate with algorithm engineers for joint optimization
Seniority
Senior, hands-on IC
