AI Infra SRE工程师（深圳/北京）
Core
Design and maintain the infrastructure for AI agents, ensuring deterministic execution, resource orchestration, and reliable model delivery for production workloads.
Role type
Senior IC AI Infrastructure SRE (Agent & Model Orchestration)
Builds
Production-ready AI agent workflows, deterministic execution layers, and scalable model inference platforms.
Domain
Cloud Native, AI Infrastructure, GPU Computing
Deliverable
production ML models
Required skills
Go, C++, Java, Rust, Linux, Kubernetes, Cloud Native, AI Programming Tools, LLM Capabilities, Agent Development, MCP, RAG, GPU Scheduling, vLLM, SGLang
Preferred skills
System debugging, AI model reliability handling, Multi-agent collaboration, Evals framework
Technologies
K8s client-go, RBAC, API Gateway, MCP Server, vLLM, SGLang, DAG, Linux
Responsibilities
Deliver single models and orchestrate workflows with idempotency and compensation; Design deterministic execution layers for high-risk agent actions; Build cluster ledgers and node lifecycle management for agent consumption; Construct platform foundations including RBAC, API gateways, and MCP servers; Implement alert noise reduction and root cause analysis for AI conclusions; Manage multi-agent collaboration and build evaluation datasets.
