DevOps Explained: The Career Behind the Curtain

Before DevOps existed as a concept, companies had two separate teams. Developers wrote code. Operations teams deployed it and kept servers running. These teams often had conflicting goals: developers wanted to ship fast, operations wanted stability, and the handoff between them was where things fell apart.

DevOps merges these responsibilities. A DevOps engineer builds systems that let developers ship code to production reliably and frequently, without manual intervention. When you push code to GitHub and it automatically gets tested, built into a Docker container, deployed to a cloud server, and monitored for errors, that entire pipeline is DevOps work.

Think of it like this. Developers build the car. DevOps engineers build the road, the traffic lights, the petrol stations, and the crash barriers. Without them, the car sits in the garage.

In Kenya, DevOps has become critical as companies move from manual server management to cloud infrastructure, from monthly releases to daily deployments, and from "it works on my machine" to "it works everywhere, reliably, at scale." Every Kenyan fintech processing M-Pesa transactions needs its production environment to be rock solid. That is a DevOps responsibility.

No two days look the same, which is part of the appeal. But a typical week for a DevOps engineer at a Nairobi fintech might include these tasks.

Monday: A developer's deployment failed because a Docker build timed out. You investigate, discover the build cache was corrupted, fix it, and update the CI/CD pipeline to prevent recurrence. Then you review a pull request from another DevOps engineer who is adding new monitoring alerts.

Tuesday: The product team wants to launch a new microservice. You set up the infrastructure: create the AWS ECS service, configure the load balancer, set up environment variables, create the CI/CD pipeline, and add monitoring dashboards in Grafana. What would have taken a week two years ago now takes a day because you have built reusable Terraform modules.

Wednesday: An alert fires at 10 AM. CPU usage on the database server spiked to 95%. You investigate, find a poorly optimised query that a developer pushed yesterday, coordinate a fix with the backend team, and add a slow-query alert to catch this earlier next time.

Thursday: You are upgrading Kubernetes to a newer version across the staging environment. This requires careful testing because a botched upgrade could take down all services. You plan the rollout, test it in staging, document the process, and schedule the production upgrade for a low-traffic window.

Friday: Infrastructure cost review. AWS bills have increased 20% this month. You identify unused resources, right-size some EC2 instances, and set up auto-scaling policies to reduce costs without affecting performance. Then you write documentation for the on-call handover to the weekend team.

The work is a blend of firefighting, building, and optimising. Some weeks are calm. Some weeks a production outage at 2 AM has you troubleshooting in your pyjamas. The unpredictability requires comfort with ambiguity and cool-headedness under pressure.

Linux. Everything runs on Linux. Servers, containers, CI/CD runners. You need to be comfortable with the command line: navigating directories, managing files, checking processes, reading logs, configuring services. If you cannot troubleshoot a failing server via SSH without a graphical interface, you are not ready for DevOps.

Docker. Containers package applications so they run the same way everywhere. Writing Dockerfiles, building images, managing container networks, and troubleshooting containerised applications are daily tasks. Docker is the entry point to container orchestration.

CI/CD platforms. GitHub Actions, GitLab CI, or Jenkins. These automate the process of testing, building, and deploying code. You write pipeline configurations that define what happens when a developer pushes code: run tests, build the app, deploy to staging, run integration tests, deploy to production.

Cloud platforms. AWS is dominant in Kenya, followed by GCP and Azure. You need to understand compute (EC2/Cloud Run), networking (VPC, load balancers), storage (S3, RDS), and identity management (IAM). Most Kenyan companies run on one primary cloud provider, and they want DevOps engineers who know it deeply.

Infrastructure as Code (IaC). Terraform or Pulumi for defining infrastructure in code files rather than clicking through cloud console UIs. This makes infrastructure reproducible, version-controlled, and reviewable. Writing good Terraform is a skill that takes months to develop.

Monitoring and observability. Prometheus for metrics, Grafana for dashboards, and tools like Datadog or New Relic for application performance monitoring. You need to set up alerts that catch problems before users do, and dashboards that show system health at a glance.

Scripting. Bash and Python for automation scripts. Many DevOps tasks involve gluing tools together with scripts: backup scripts, deployment scripts, cleanup scripts, migration scripts. You do not need to be a software engineer, but you need to write reliable scripts.

DevOps is rarely a first career. Most DevOps engineers arrived via one of three routes.

Route 1: Backend developer to DevOps. The most common path. You are a backend developer who gets frustrated waiting for someone else to deploy your code, so you learn Docker, set up a CI/CD pipeline, and gradually take on more infrastructure work. One day you realise you spend more time on infrastructure than application code, and you make the switch official. This path works because you already understand how applications are built.

Route 2: System administrator to DevOps. You managed Linux servers, maybe at a Kenyan ISP, hosting company, or corporate IT department. You already know networking, server management, and troubleshooting. Learning containerisation, cloud platforms, and CI/CD bridges the gap to DevOps. This path works because you already understand operations.

Route 3: Deliberate entry. Less common but possible. Learn Linux administration, Docker, and AWS simultaneously through structured courses and hands-on projects. Build a homelab (even on a cheap laptop running VirtualBox). Set up a CI/CD pipeline for a personal project. Deploy a web application to AWS with monitoring. Document everything as portfolio pieces.

Whichever route you take, the key is hands-on practice. Reading about Docker does not make you a Docker user. Running 100 containers, breaking things, and fixing them does. Set up a personal project with a full deployment pipeline: GitHub repo, GitHub Actions CI/CD, Docker containers, AWS deployment, and Prometheus monitoring. That project alone demonstrates more than any certification.

McTaba's 26-week programme includes deployment training because every developer should understand how their code gets to production. If DevOps interests you, that foundation plus additional study of cloud infrastructure and monitoring tools builds a strong starting point.

Junior DevOps / Junior Cloud Engineer (0-2 years): KES 60,000 to KES 150,000. These roles exist but are uncommon because most companies want experienced DevOps professionals. Junior positions tend to appear at larger companies that can afford to train, or at companies where the role is combined with backend development.

Mid-Level DevOps Engineer (2-4 years): KES 150,000 to KES 350,000. At this level, you manage production infrastructure independently, handle incidents, and build CI/CD pipelines. Most Kenyan companies hiring DevOps engineers want someone at this level or above.

Senior DevOps / SRE (4-7 years): KES 350,000 to KES 700,000. You architect infrastructure for the company, lead migration projects, mentor juniors, and define reliability standards. Banks, telecoms, and well-funded fintechs pay at the top of this range.

Principal / Head of Infrastructure (7+ years): KES 500,000 to KES 1,000,000+. These are rare roles at large companies. You define the entire infrastructure strategy and manage teams of DevOps engineers.

Remote DevOps roles are particularly lucrative. International companies hire Kenyan DevOps engineers for USD 3,000 to USD 8,000 monthly because infrastructure management does not require physical presence. The time zone overlap with Europe makes remote DevOps work practical for Kenyan professionals.

Cloud certifications boost earning potential in this field more than in most other tech roles. AWS Solutions Architect, AWS DevOps Engineer Professional, and Certified Kubernetes Administrator (CKA) are the most valued certifications in Kenya's market.

Several trends are shaping the DevOps landscape in Kenya.

Cloud migration is accelerating. Kenyan companies that ran their own servers are moving to AWS and GCP. This creates projects, and projects need DevOps engineers. Banks that previously insisted on on-premise infrastructure are now deploying hybrid architectures with cloud components.

Platform engineering is emerging. Instead of every team managing their own infrastructure, companies are building internal developer platforms: self-service systems where developers deploy without needing to understand Kubernetes. Platform engineers build these systems, and it is a natural evolution for senior DevOps professionals.

AI is changing operations. AI tools are starting to help with incident response, log analysis, and infrastructure optimisation. This does not replace DevOps engineers. It changes what they spend their time on. Instead of manually analysing logs, they configure AI assistants to surface anomalies. The role shifts toward higher-level architecture and strategy.

Security is merging with DevOps. "DevSecOps" integrates security practices into the deployment pipeline: automated vulnerability scanning, dependency checking, and compliance enforcement. As Kenyan regulatory requirements tighten, especially in financial services, DevOps engineers who understand security earn even more.

The bottom line: DevOps demand in Kenya will keep growing. Every new digital product needs infrastructure. Every company moving to the cloud needs someone to manage it. And the supply of qualified DevOps engineers in Kenya remains far below demand.