I’ve been burned by workflow migration promises before, so I’m asking this carefully: how much of AI-generated workflows actually makes it to production without major editing?
We’re evaluating a move away from our n8n self-hosted setup, and one of the selling points I keep hearing is that you can describe what you want in plain language and the platform generates workflows automatically. That sounds amazing, but I’m skeptical about how much actually works out of the box.
The reason I’m asking is that we’ve got maybe 40 active workflows running on n8n right now. Some are simple, some have branching logic, error handling, custom fields we’ve built over time. The idea of describing each one to an AI and getting something production-ready sounds efficient, but I’d rather hear from someone who’s actually tried it than buy into the marketing version.
What’s been your experience? When you ask an AI copilot to generate a workflow from a plain-English description, how much tweaking do you actually need to do before it’s ready for production? Is this actually faster than rebuilding manually, or is it just moving the effort around?