Devops Engineer (Bengaluru)

Devops Engineer (Bengaluru)

15 Nov
|
Xoriant
|
Bengaluru

15 Nov

Xoriant

Bengaluru

Job Requirements

Cisco's AI Platform Team builds the infrastructure behind NeuralFabric: enterprise AI model training, evaluation, and deployment running across GPU clusters in AWS, GCP, and Cisco IT. We are bringing on contract capacity so our infrastructure engineers can stay focused on scaling the platform while day-to-day reliability, automation, and operational tooling get owned properly.

Your Impact

- Operate multi-cluster GPU infrastructure. Keep GPU clusters across AWS, GCP, and Cisco IT healthy and efficiently utilized as capacity scales from tens into the hundreds.
- Run the reliability practice. Define SLIs, SLOs, and error budgets for training, evaluation, and inference services, then hold the platform to them.
- Own incident response end to end. Take on-call shifts, act as incident commander during outages, and run blameless postmortems that close with committed fixes rather than open action items.
- Cut toil and alert noise. Replace manual operational work with automation, and tune alerting so every page is actionable and every actionable failure pages.
- Automate provisioning and deployment.

Extend

Terraform-based infrastructure-as-code and CI/CD pipelines so cluster setup, upgrades, and lifecycle management stop being manual.

- Keep distributed AI workloads running. Troubleshoot multi-node training and inference issues spanning GPU drivers, CUDA runtimes, Ray, container images, and storage or network bandwidth.
- Plan capacity and improve cost efficiency. Forecast GPU demand, then apply Spot instances, fractional GPU configurations, autoscaling, and scheduling policy to raise utilization and reduce spend.
- Hold the security line. Maintain multi-tenant GPU isolation, namespace policies,



access control, and secrets handling; no credentials in source code, images, or IaC.
- Document and hand off. Leave runbooks and automation the permanent team can own after the engagement.

Minimum Qualifications

- Production on-call experience in a real rotation, with incident command and blameless postmortem practice.
- Production Kubernetes and container experience (Docker), with cloud-native infrastructure patterns.
- Hands-on production ownership on at least one major cloud (AWS, GCP, or Azure).
- Terraform or OpenTofu proficiency.
- Observability depth with Prometheus, Grafana, or equivalent for metrics, logging, and alerting, including dashboard and alert design.
- Solid automation skills in Python, Bash, or Go.
- Networking fundamentals: VPCs, load balancers, DNS, firewalls, cross-cloud connectivity.
- CI/CD experience with GitHub Actions, GitLab CI, Jenkins, or ArgoCD.
- Proven ability to troubleshoot complex distributed systems, largely self-directed.

Preferred Qualifications

- GPU infrastructure and AI/ML workloads: Ray, Kubeflow, MLflow, or similar.
- NVIDIA GPU orchestration: A100/H100 configuration, driver and CUDA runtime management.
- Distributed training networking: RDMA, InfiniBand, EFA, NCCL.
- Distributed tracing and OpenTelemetry instrumentation across services.
- Progressive delivery: canary and blue/green rollouts with automated rollback.
- Chaos or fault-injection testing, game days, and disaster-recovery drills.
- Multi-cloud networking, unified storage abstractions, and disaster recovery.
- FinOps and cost optimization: Spot, Reserved Instances, Savings Plans.
- Establishing an SRE function where one did not previously exist. .

📌 Devops Engineer (Bengaluru)
🏢 Xoriant
📍 Bengaluru

Reply to this offer

Impress this employer describing Your skills and abilities, fill out the form below and leave Your personal touch in the presentation letter.

Subscribe to this job alert:

Get the latest job offers by email for: devops engineer (bengaluru) / bengaluru

Subscribe to this job alert:

Get the latest job offers by email for: devops engineer (bengaluru) / bengaluru