DevOps Engineer

Nomiso is scaling an on-premise Agentic AI Platform from proof-of-value into enterprise production for tier-1 telecom operator. As a Devops Engineer, You will own the foundation that powers self-hosted LLM endpoint (vLLM), Kubernetes infrastructure, and automated GitOps Pipelines within a secure client environment.

What you will do

• Build and operate the Kubernetes platform on Linux VMs inside the client network: cluster setup, upgrades, networking, storage and node management.
• Own the CI/CD pipelines: Docker image builds, security scanning, Helm packaging and GitOps deployment through ArgoCD across dev, staging and production.
• Manage environments as code with Terraform or Ansible, including Linux hardening, patching and configuration drift control.
• Run the data and messaging layer the application depends on: SQL Server connectivity, Redis, Kafka and object storage, working with the client DBA team where they own the service.
• Deploy and operate the self hosted LLM serving stack (vLLM or an OpenAI compatible gateway)
• and ML tooling such as MLflow, including GPU node scheduling where available.
• Set up observability end to end: Prometheus, Grafana, Loki and OpenTelemetry tracing per ticket,
• with alerts routed to Teams and the client ops tooling.
• Manage secrets, certificates and service identities through Vault and Active Directory, and keep the platform aligned with client security and VPN access controls.
• Define SLOs, run capacity and latency reviews for the integration path (LLM endpoint, ACS, NMS, inventory APIs) and tune resources to keep per ticket decisions within the agreed window.
• Write runbooks, handle incidents, lead root cause reviews and train client operations staff ahead of handover.

What you bring

• 4 to 5 years in DevOps or platform engineering, with strong Linux administration (RHEL, Ubuntu or similar): systemd, networking, storage, performance troubleshooting.
• Production Kubernetes experience: Helm, ingress, RBAC, persistent volumes, resource limits, cluster upgrades.
• Docker and container security basics, including image scanning and non root builds.
• CI/CD with GitHub Actions, GitLab CI or Azure DevOps, and GitOps with ArgoCD or Flux.
• Infrastructure as code with Terraform and configuration management with Ansible.
• Scripting in Bash and Python for automation and operational tooling.
• Observability stack experience: Prometheus, Grafana, Loki or ELK, OpenTelemetry.
• Comfortable in an on premise environment with VPN access, firewalls, proxies and Windows Active Directory integration.
• Clear written communication: runbooks, change notes and incident summaries that client teams can follow.

Good to have

• Running LLM or ML workloads on Kubernetes: vLLM, GPU scheduling, MLflow, model serving.
• Exposure to SQL Server, Redis and Kafka operations.
• Rancher, Azure Stack HCI or other on premise Kubernetes distributions.
• HashiCorp Vault or CyberArk, certificate automation.
• Telecom or large enterprise client delivery experience.
• Certifications such as CKA, RHCSA or Terraform Associate.

How we work

Small team, direct access to the technical lead, measured delivery in phases with the client. You will be the single owner of the platform, so judgement, documentation and calm incident handling matter as much as tool knowledge.

Hybrid working from Bengaluru or Hyderabad, with no travel required.