微信视频号-大模型ai infra高级工程师
Core
Develop model training and inference platforms supporting real-time training/inference of ultra-large sparse models and high-concurrency, low-latency execution for recommendation systems.
Role type
Senior IC AI Infrastructure Engineer (Model Training & Inference)
Builds
Model training platform, model inference platform, recommendation system infrastructure
Domain
AI/ML Infrastructure, Recommendation Systems
Deliverable
production ML models
Required skills
C/C++, Linux development, Data structures and algorithms, Python/Shell scripting, Deep learning frameworks (PyTorch/TensorFlow), GPU optimization, Mathematical/algorithmic foundations
Preferred skills
GPU optimization experience, Open-source framework contributions
Technologies
PyTorch, TensorFlow, Linux, C/C++, Python, Shell, GPU
Responsibilities
Develop model training platforms for real-time training of ultra-large sparse models; Develop model inference platforms for high-concurrency, low-latency execution; Support efficient iteration of large-scale recommendation business matrices (short video, live streaming, red dot, text/image recommendations).
