Staff AI-Native DevOps Engineer

Apply

Workplace Type

Remote

Service Management, Enterprise Cloud, Custom App Development, Platform, ITSM, Security Operations, Customer Service Management, ITOM, GRC, HR Service Delivery, and digitize workflows

DevOps Engineer FAQ

Common questions from people exploring this role.

What is the difference between DevOps and a Sysadmin role?

Do I need to know how to code to be a DevOps Engineer?

Is Kubernetes required for every DevOps role?

How does DevOps differ from Site Reliability Engineering (SRE)?

The Conversational AI team at ServiceNow builds the core platform behind ServiceNow Otto, our conversational AI experience. We are a group of applied scientists and software engineers working at the intersection of research and production. Our work spans the full stack that makes Otto capable: agent orchestration, including harness engineering, prompt tuning, agent skills, policy design, and quality evaluation across languages; the A2A and MCP protocols that let agents collaborate and connect to tools; background channels and conversational triggers that bring agents into action; AI-native licensing and pricing; and the central caching layer that connects the ServiceNow platform to the services powering Otto.

We are seeking a Staff AI-Native DevOps Engineer to join a new DevOps practice within the Conversational AI team. You will own the build, deployment, and reliability of the systems behind ServiceNow Otto, our conversational AI platform, from commit through production across our SaaS environments. Your work will span CI/CD pipelines, release automation, infrastructure on Kubernetes, observability, and the operational tooling that lets agent orchestration, MCP and A2A services, and the central caching layer ship safely and frequently. You will partner closely with applied scientists and software engineers to make AI-native delivery repeatable, measurable, and secure. This role combines deep expertise in cloud infrastructure, containerization, orchestration, and infrastructure as code with a practical understanding of how ML and LLM-based services behave in production, and a bias toward pragmatic engineering with measurable impact on deployment velocity and service reliability.