In this blog, I take a real Power Query data cleaning problem and try to solve it using AI. The idea is simple. Can AI handle the problem the same way a human would, and where does it start to struggle? I break the solution into three parts. First, I use a very basic AI prompt and show why the result works but doesn’t scale. Then I improve the prompt and test a more flexible AI-generated solution, while pointing out some hidden issues that can appear with larger data. Finally, I walk through my own Power Query approach and explain the thinking behind it. The focus is on building scalable logic, reducing hard-coding, and understanding how data is structured before writing M code. This video is useful if you use AI with Power Query and want to understand when to trust it and when to rely on your own data thinking. I’ve also shared the full solution and AI prompts as a downloadable file linked below.
I competed with AI and Lost but…
