推荐大模型训练/推理优化工程师-电商
Core
Developing and optimizing LLM training and inference technologies specifically for recommendation systems to solve engineering bottlenecks and reduce costs.
Role type
Senior IC LLM training and inference optimization engineer
Builds
High-performance recommendation system architectures and optimized LLM inference frameworks
Domain
E-commerce / Recommendation Systems / Large Language Models
Deliverable
production ML models
Required skills
C/C++, Python, CUDA development, PyTorch, TensorFlow, model quantization, graph compilation, operator optimization, KVCache, PD separation, Megatron, Deepspeed, VLLM, SGLang, TRT-LLM
Preferred skills
Recommendation/ad/search model optimization experience, contributions to LLM open-source communities
Technologies
TensorRT, Triton, Cutlass
Responsibilities
Research and implement LLM training/inference solutions for recommendation scenarios; Provide high-automation, high-performance model optimization for frameworks; Design and implement next-gen recommendation system architectures using open-source LLM components; Collaborate with algorithm teams on end-to-end generative recommendation innovation.