Am I the only one who thinks AI hiring tools aren't ready for real use?
Sat in on a demo in Austin last month where the AI screening tool rejected a candidate with 12 years experience because his resume listed 'PowerShell' instead of 'Power Shell' and we lost him to a competitor. Has anyone actually seen one of these tools handle edge cases without a human babysitting it?
Walk me through what happened after the tool flagged him. Did the recruiter catch it in review, or did the rejection go out automatically before anyone looked? That part matters more than the spelling glitch itself. If a human saw the flag and still let it go, that's a process problem, not a tech problem. Curious how much of the demo was scripted versus them letting you throw messy real resumes at it.
I used to figure the tech was the whole story, but you're right that it comes down to whether a person actually looked before hitting send. In my experience the demo always uses clean samples, so the real test is what happens when you hand it a stack of messy ones and watch who reviews the flags.