I keep seeing platforms claiming they can turn plain language descriptions into ready-to-run workflows using AI. That sounds incredible if it’s real. But I’m skeptical.
From what I’ve seen with other automation tools, AI-generated code or workflows require a lot of refinement before they’re actually production-ready. I’m wondering if AI copilot workflow generation is different, or if it’s the same story—you describe what you want, the system generates something in the ballpark, and you spend days cleaning it up and handling edge cases.
If it actually works, that would meaningfully change the economics of automation. You could onboard less technical people into workflow building, reduce time-to-value, and keep costs down. But if you’re spending as much time debugging AI-generated workflows as you would building them manually, the ROI proposition falls apart.
For anyone who’s tried this, what was the reality? How many AI-generated workflows actually made it to production without a complete rebuild?