Senior Vibe Coder
In the past I've mentioned that I used AI to refresh this site, in everything from migrating the data from the previous platform to a new one, to choosing a new platform to use. That's far from the first time I have used AI; in fact, I use it for work more than anything else. I have been using it for everything from single code functions to full build and deployment scripts, which are now faster than ever before. Even the image in the banner for this entry was generated with Google Gemini.
Indeed, the past year has been a somewhat annoying, somewhat entertaining year where people swung from predicting that "AI is going to replace developers" to the rise of a new job position: the vibe code cleanup specialist.
Understanding the "Vibe code cleanup specialist"
Tapan Patel provides an insightful definition of what a "vibe code cleanup specialist" is:
A vibe coding cleanup specialist is a senior software engineer who specializes in fixing AI-generated code and transforming vibe-coded prototypes into secure, scalable, production-ready applications.
Tapan Patel - Vibe Coding Cleanup Specialist | Everything You Need to Know in 2025
Indeed, you can see that many people have immediately embraced this title based on their LinkedIn profiles:
And while for some this might be tongue-in-cheek or, perhaps, just plain desperate considering the volatile software development market, the software development community is at odds with what to make of it:
The emergence of vibe coding cleanup specialists has stirred debate in the tech community. Some developers dismiss it as "AI babysitting," arguing that senior engineers shouldn't spend time cleaning up messes that AI creates.
Others embrace the role, noting it allows them to focus on high-level architecture and system design rather than writing boilerplate code from scratch. The reality is that vibe coding excels at speed but fails at the details that matter for production systems.
Tapan Patel - Vibe Coding Cleanup Specialist | Everything You Need to Know in 2025
Complicated impact
The impact of AI and vibe coding is being felt most in hiring processes. My employer currently faces a problem where take-home coding exams -- excellent ways of screening candidates in the past -- are now fed through AI agents that churn complete, industry-standard compliant answers. While one could argue that the easiest way to mitigate this is through live coding exercises or direct interviews, this exam was explicitly designed to remove that need. Because this step was made explicitly to help managers hire for their teams without having to talk to all applicants, they've needed to spend more time on the hiring process rather than on their regular tasks.
On the other hand, senior developers (like myself) have been using AI to leverage our own output. The ability to invoke AI assistance within Visual Studio (e.g., Github Copilot) has made it much easier to write code faster in the context of already-existing legacy code. To the chagrin of TDD purists, AI has also been very helpful in generating unit tests, removing the tedium of writing out repetitive scenarios, and surfacing edge cases that do not have unit tests. In other cases, it acts as a guide to quickly code in unfamiliar tech stacks or programming languages -- whether it be UI frameworks like React or infrastructure definitions written with YAML.
In some ways AI delivers the promise of Rapid Application Development -- a term from the 1990s -- into practical reality. For people who know what they are doing, AI is a Godsend. For senior, "10x" developers, it multiplies that capability some more. But the drawbacks are mirrored -- for junior developers dependent on AI, bugs introduced with copy-pasted code will be 10x harder to find, owing to both unfamiliarity with the code and the algorithms used. The sword isn't just double-edged -- it's poison-tipped on both sides.
The Future
I do not think senior software developers have much to fear from AI agents, vibe coders, or "vibe code cleanup specialists". In many ways technical skills nurtured through waves of hype combined with actual experience working with customers will remain much more valuable in the near future. It's merely time for them to embrace the title of "senior vibe coder".
However, more discretion is advised for junior developers; while the temptation to use AI exclusively to accomplish tasks is prevalent, it is prudent for developers with less experience to have the AI explain its code line-by-line. It shouldn't be a choice of whether or not to use AI, but rather an opportunity to use AI to help developers learn what they are actually doing. This is easier said than done, but skipping the learning part is exactly what makes a dev vulnerable to being replaced by AI entirely.