I don't know what the secret is, because "smartness" is vague enough that it's not objectively measurable. We aren't talking about lifting weights or running th

I don't know what the secret is, because "smartness" is vague enough that it's not objectively measurable. We aren't talking about lifting weights or running the mile where you can know your own abilities and compare them to another person's - intelligence just doesn't work that way. Interviews and conversations are the best way I know of to tell how "smart" someone is, but there are problems there as well. The problem I have in mind (there are others) is that it's not very easy to tell if someone is smarter than you are. I find that it's easy to tell if someone is less intelligent than I am because they don't recognize their assumptions, draw incorrect inferences or are otherwise sloppy. However, in a sense I'm just testing for the absence of stupidity and not the presence of intelligence. I find it very difficult to tell if someone's actually smarter than I am or just more knowledgeable. What I would suggest is that you have several people interview this person separately (ideally on separate topics) and use their opinions to formulate a judgement. If, after conversations on different topics, 5 engineers say "this girl seems to know her stuff" then it's a much safer bet that this is person is a bar-raiser than if it's just you. Yes, you'll likely be turning lots of people away, but that's what happens if you keep raising the bar.