DevOps Engineer Job: Day-to-Day Work Inside a Company

This course contains the use of artificial intelligence. DevOps Engineer Job: Day-to-Day Work Inside a Company is a practical course on how a DevOps engineer works inside a real engineering organization. You will learn what a DevOps engineer owns on the path from code change to production, who you work with, how your day and week are organized, how CI/CD pipelines, GitOps, Kubernetes, deployment strategies and infrastructure as code fit together, how a release is promoted, verified and handed over, and what to do when a production deployment goes wrong.

The course follows a new DevOps engineer through the first weeks on the Developer Delivery Platform team at Walnut Keystone Corporation, a fictional global company that sells furniture online. The company and its people are fictional; the tools and ways of working are real and current. GitHub Actions, GitHub Container registry, Amazon EKS, Helm, Argo CD, Terraform, LaunchDarkly, Datadog and PagerDuty run the story, and current alternatives are named neutrally for CI/CD, GitOps, managed Kubernetes, infrastructure as code, feature flags and observability. Work items, pull requests, a change record and an incident are followed step by step, with their reviews, approvals and evidence. Approvals, change freezes and the manual production sync are Walnut Keystone policy, and the course shows where these practices vary between companies.

The DevOps engineer role

• How one code change travels from pull request through CI, a container image, GitOps configuration and a Kubernetes deployment to verified production
• Where DevOps ownership starts and stops on the software delivery path
• How DevOps differs from site reliability engineering (SRE), platform engineering, cloud engineering and release management, and why titles vary between companies

Your team and how to reach people

• What a Developer Delivery Platform team owns: reusable CI/CD workflows, GitOps delivery patterns, release tooling, documentation, guardrails and developer support
• What product engineering, SRE, cloud, security and release management partners own, and the evidence each one needs from you
• How help requests move from a Slack channel to a Jira work item, and how to escalate by the failing layer

The workday, the week and the support rotation

• A DevOps workday: board and handoff review, pipeline work, code review, developer support, releases and the end-of-day handoff
• Delivery planning, delivery sync, release readiness, platform office hours and the retrospective, and how to prepare for each
• The delivery-support rotation with a primary and a secondary, shadowing before your first supervised week, and handoffs across time zones
• Standard, normal and emergency changes, Git-driven production access, break-glass access and release freezes

Core DevOps work

• CI pipelines in GitHub Actions: jobs, steps, runners, artifacts, security scanning, failure diagnosis and faster feedback
• Continuous delivery and GitOps with Helm charts, a config repository, Argo CD sync and health, and Flux as an alternative
• Rolling updates, blue-green and canary deployments, and feature flags with LaunchDarkly or Unleash
• Infrastructure as code with Terraform and OpenTofu, plan review, least privilege and GitHub Actions OIDC with AWS in place of long-lived cloud credentials

One release workflow upgrade from start to production

• Intake in a help channel, current-state mapping, design review and a reusable GitHub Actions workflow pull request
• Adoption by the product team, image publication to GitHub Container registry, a GitOps staging promotion and staging validation
• Release readiness, the production config pull request, a manual Argo CD sync, Datadog verification, handover and runbook updates

Tools, environments and observability

• GitHub Actions and current alternatives: GitLab CI/CD, Jenkins Pipeline, Azure Pipelines and CircleCI
• Container images, registries, Kubernetes Deployments and Pods, namespaces, Helm charts and managed Kubernetes on Amazon EKS, Azure Kubernetes Service and Google Kubernetes Engine
• Release observability with Datadog deployment tracking, Prometheus metrics, Grafana dashboards, alerts and PagerDuty on-call routing

The language of the job

• CI and CD vocabulary: continuous integration, continuous delivery and continuous deployment, pipeline, job, step, runner, artifact and environment
• Kubernetes and GitOps vocabulary: container image, registry, Pod, Deployment, namespace, Helm chart, desired state, sync, health, drift and reconciliation
• Release vocabulary: rollout, rollback, feature flag, release window, freeze, change record and the current DORA metrics, including change lead time, change fail rate and deployment rework rate

When things go wrong

• A failed production deployment minute by minute: Datadog monitor, PagerDuty page, SRE incident lead, release hold, GitOps revert pull request, Argo CD sync, verification and resolution
• Who leads, who owns the application and who owns the rollback mechanics during a release incident
• A blameless postmortem: the incident timeline, contributing conditions, what worked and owned delivery-system follow-up actions

Before day one and the interview

• A DevOps skills checklist: Git, the Linux shell, networking basics, YAML, CI/CD, containers, Kubernetes, infrastructure as code, observability and troubleshooting evidence
• Your first 90 days as a DevOps engineer: system mapping, shadowing support, a small reviewed change, supervised delivery work and partner trust
• DevOps interview tips for delivery-system, troubleshooting, Kubernetes and deployment-safety questions, experience stories, pipeline and infrastructure-as-code practical exercises, and questions to ask the team

Practical tools you can use at work

Apply each part of the job with downloadable guides, templates and workbooks: the Team Contact Map for DevOps Engineers, the Meetings and Rotations Guide for DevOps Engineers, DevOps Work Templates, the DevOps Job Vocabulary, the Incident Card for DevOps Engineers, the DevOps Engineer Skills Self-Check and 90-Day Plan, the DevOps Engineer Interview Preparation Guide, and the DevOps Delivery Tool Map.