Gosh, that’s one more pretty good red flag to look for in interviews, thanks for the tip! [edit, to tame the sarcasm in my comment] If your project is very la

Gosh, that’s one more pretty good red flag to look for in interviews, thanks for the tip! [edit, to tame the sarcasm in my comment] If your project is very large and feature estimation is an issue, that’s because both (the feature actual scope and its estimation) are still in the discovery phase. So that means that the discovery/design/estimation is a feedback loop, not a path. If your team/project structure or your customer can’t accept/manage that, that’s the issue; not the estimation in itself. Independently, if an AI can satisfyingly and consistently estimate the delay/complexity of a user story into a working feature within a given product, by a given team, it may already be near cutting/splitting it into smaller ones. And if that still makes a working feat/product for humans in the end (that is, something significantly better than what Concur or SAP can be for instance), your AI may even be in the range where it can implement it. Both hunches that would make the hiring company a hard pass for me, in that either it’s not planning/organised in a sane way, either my role is not needed.