Senior DevOps / SRE Engineer (AWS & GPU Infrastructure)
Description
About the Role
We are looking for a Senior DevOps / SRE Engineer to own the operations of our AWS cloud platform and core production systems, while gradually supporting our AMD and NVIDIA GPU clusters. You will help keep our services reliable and continuously improve our automation, monitoring, alerting, and incident response capabilities.
Key Responsibilities
AWS Cloud Platform & Production Systems
- Manage day-to-day operations of AWS resources and core production systems, ensuring reliability, availability, and security.
- Maintain cloud networking, load balancing, storage, and access control, while optimizing resource allocation and cloud costs.
- Strengthen backup and recovery, capacity planning, and disaster recovery plans.
Kubernetes & Container Platform
- Operate and maintain Kubernetes clusters and containerized environments, including nodes, scheduling, networking, and storage components.
- Improve deployment, release, rollback, and autoscaling workflows for applications and inference services; maintain CI/CD pipelines.
- Partner with engineering teams to resolve runtime and environment compatibility issues.
GPU Cluster Operations (gradually ramping up)
- Support day-to-day maintenance and troubleshooting of AMD Instinct and NVIDIA GPU clusters.
- Maintain GPU drivers, ROCm / CUDA, container runtimes, and Kubernetes GPU-related components.
- Diagnose GPU issues such as card drop-offs, memory errors, and unavailable devices, and coordinate with vendors on repairs.
Monitoring, Alerting & Incident Response
- Build and maintain monitoring, logging, and alerting systems covering applications, cloud resources, and GPUs.
- Participate in the on-call rotation and respond to production incidents, identifying issues across network, node, container, and service layers.
- Lead post-incident reviews and drive long-term fixes for recurring problems.
Operational Excellence & Automation
- Create runbooks, inspection checklists, and change management processes.
- Drive infrastructure automation using scripting and tools such as Terraform and Ansible.
Job requirements
- 5+ years of experience in DevOps, SRE, or infrastructure operations, including production system maintenance and incident handling.
- Hands-on experience operating AWS in production, with solid knowledge of compute, networking, storage, IAM, and monitoring services.
- Strong experience with Kubernetes and containers (Docker / containerd), including cluster maintenance and troubleshooting.
- Strong Linux fundamentals, with the ability to troubleshoot system, network, disk, and process-level issues.
- Experience operating GPU clusters (either AMD or NVIDIA).
- Willingness to participate in an on-call rotation (Asia time zones).
- Proficiency in at least one scripting or programming language: Bash, Python, or Go.
Nice-to-have qualifications
- Familiarity with AMD ROCm, or inference frameworks such as vLLM / SGLang.
- Experience with monitoring tools such as Prometheus, Grafana, and CloudWatch.
- Experience with Terraform, Ansible, or GitOps practices.
- Knowledge of InfiniBand / RoCE / RDMA and multi-node, multi-GPU communication troubleshooting.
- Experience using AI / agent tools to improve operational workflows.
- Ability to communicate with overseas vendors in English
Scope & Collaboration
This role focuses on cloud platform and production service operations, with GPU infrastructure as a growing area of responsibility. You will work closely with engineering and ML teams to support deployment and runtime reliability. Model training, algorithm research, and inference framework development are owned by the respective engineering teams.
Salary: Negotiable based on qualifications and experience.
Headcount: 1; referrals are highly welcomed.
Contact information
MAIL:karen.hsiao@tron.network
Skills
Sign in to read and join the discussion.