Senior DevOps Engineer

Role Overview

Key Responsibilities

• Design, build, and maintain AWS, Google Cloud Platform, or Microsoft Azure infrastructure using Terraform, Pulumi, or CloudFormation.
• Architect and manage CI/CD pipelines for applications and ML model deployment.
• Build and maintain Kubernetes clusters and manage containerized services and ML inference workloads.
• Own logging, metrics, tracing, and alerting using tools such as Prometheus, Grafana, ELK/EFK, and Datadog.
• Support feature stores, model registries, training pipelines, and GPU cluster management.
• Manage GPU provisioning and optimize costs using spot instances, autoscaling, and reserved capacity.
• Drive IAM, secrets management, network policies, and SOC2/ISO compliance practices.
• Lead incident response, on-call rotations, postmortems, and SRE culture using SLOs and SLIs.
• Mentor junior engineers and collaborate with data science and ML teams.
• Manage multi-region and multi-cloud deployments, disaster recovery, capacity planning, and FinOps governance.

Required Skills and Experience

• Deep expertise in one hyperscaler and working knowledge of a second among AWS, Google Cloud Platform, and Microsoft Azure.
• Knowledge of compute, networking, VPCs, load balancers, CDNs, storage, and IAM.
• Infrastructure-as-code and automation experience with Terraform, Pulumi, CloudFormation, Ansible, Chef, Puppet, or GitOps tools.
• Container and orchestration experience with Docker, Kubernetes, Helm, and Kustomize.
• CI/CD experience with Jenkins, GitLab CI, GitHub Actions, ArgoCD, or CircleCI.
• Observability experience with Prometheus, Grafana, Alertmanager, ELK/EFK, Loki, Datadog, New Relic, Jaeger, or OpenTelemetry.
• Security knowledge including Vault or Secrets Manager, IAM policies, network security, container security, and compliance frameworks.
• Strong Python skills for automation and ML tooling integration.
• SQL/NoSQL operations and messaging experience with Postgres, MongoDB, Redis, Kafka, RabbitMQ, or Pub/Sub.

Preferred Qualifications

• AWS, Google Cloud Platform, or Microsoft Azure Solutions Architect certification, or CKA/CKAD certification, is preferred.
• Experience operating LLM inference infrastructure at scale is preferred.
• Experience optimizing costs for large-scale GPU training is preferred.
• Open-source infrastructure or MLOps tooling contributions are preferred.
• BE/BTech in Information Technology or a related discipline is required.