CareerPlanSign in

AI大模型架构师

Shenzhen, China💼 Full-time🗓 2026-09-28

Core

Designing and optimizing the technical architecture for large-scale LLM/VLM/MoE models, covering training, inference, and infrastructure.

Role type

Senior IC AI Large Model Architect

Builds

Production-scale large language models and AI infrastructure

Domain

Artificial Intelligence / Large Language Models

Deliverable

production ML models

Required skills

Large model training engineering, Distributed training systems (3D parallelism, ZeRO), Model optimization (MoE, MLA), Post-training algorithms (SFT, DPO, GPRO), Inference deployment (vLLM, KV Cache), AI infrastructure planning

Preferred skills

Multi-card cluster training experience, Reinforcement learning experience

Technologies

PyTorch, Megatron, DeepSpeed, vLLM

Responsibilities

Planning overall technical architecture for LLM/VLM/MoE models; Optimizing large-scale distributed training systems; Planning AI infrastructure including chips, storage, and communication systems.

Sourced via tencent · Listed on CareerPlan, which tracks 844,000+ jobs from 20+ sources.