Agent Infra SRE工程师(Agent治理方向)(深圳/北京)
Core
Ensuring service availability for AgentRuntime across domestic and international regions, including on-call duties, alert response, incident management, and observability system construction.
Role type
Senior Site Reliability Engineer (Infrastructure)
Builds
AgentRuntime platform, multi-region sandbox resource pools, and SRE toolchains.
Domain
Cloud Infrastructure / Distributed Systems / Observability
Deliverable
infrastructure
Required skills
Kubernetes administration, Go or Python programming, MySQL operations, observability stack (Prometheus/Grafana), incident management, capacity planning, CI/CD pipeline management.
Preferred skills
TKE or self-hosted K8s platform experience, SLO/SLI体系建设 experience, sandbox resource management.
Technologies
Kubernetes, OpenTelemetry, Prometheus, CLS, MySQL, Go, Python, Shell, CI/CD.
Responsibilities
Monitor and maintain multi-region sandbox resource pools (CPU/memory/storage) and optimize elastic scaling strategies; Build and maintain observability systems covering metrics, tracing, and logging for sandbox creation, scheduling, networking, and storage; Lead change management and release stability through canary deployments, rollback mechanisms, and CI/CD integration; Develop SRE toolchains including self-healing scripts, inspection tools, and capacity prediction models; Participate in platform reliability design to resolve bottlenecks like CDB slow queries and sandbox startup latency.
Seniority
Senior, hands-on IC