I started as a DevOps intern and have now joined my company full time. Most of the Bash and Python automation I need is generated with AI because my team prioritizes completing tasks quickly. As a beginner, I sometimes feel inadequate because I cannot yet write large automation scripts from scratch.
How deeply should I learn Bash and Python so I can understand, debug, modify, and eventually write these scripts myself? I am planning to look for a new role around May or June next year, so I would also appreciate advice on what skills and projects I should focus on to become a stronger candidate for a well-paying DevOps position.
My current experience includes beginner-to-intermediate AWS, some Azure, Docker, Linux, Kubernetes, Terraform, and Jenkins. I would like to keep using AI productively without becoming dependent on it or missing important fundamentals.
5 Answers
For DevOps, become comfortable with Bash because operational work happens heavily in terminals and existing shell scripts are common. You should also learn enough Python to write and maintain straightforward automation; you do not need to become a full-time Python software engineer. Python is useful for larger scripts, reusable functions, APIs, and working with existing packages.
You already have a broad toolset with cloud platforms, Linux, containers, Kubernetes, Terraform, and Jenkins. Rather than immediately adding more technologies, go deeper with what you know. Build and troubleshoot complete projects, automate deployments, handle failures, and be able to explain the design decisions. For interviews, real examples of incidents you diagnosed and fixed will usually be more valuable than claiming fluency in many tools.
A useful test is to take an AI-generated script you used recently and explain every line without looking anything up. If you cannot, that shows exactly what to study. The goal is not to ban AI; it is to make sure you can still work when the generated code is incomplete, subtly incorrect, or unsuitable for your environment.
A good habit is to treat the first AI-generated script as a draft rather than a finished solution. Read every section, test failure cases, add appropriate error handling, and deliberately change or extend the logic. Ask what happens if a command fails halfway through or the input is malformed. That process will build the skills you need while still letting you work efficiently.
Yes, you should learn both, especially well enough to understand AI-generated code and handle interview questions. You do not need to stop using AI, but you should be able to read a script, explain it, debug it, change its behavior, and tell when something is wrong. That practical understanding matters more than personally typing every line.
Even teams that use Python heavily still have older or quick-and-dirty shell scripts in production. You do not need to love Bash, but you should be able to read, troubleshoot, and safely modify it.