“She took someone else's code. You know it, she knows it, so the conversation is done.”

“That code gap tells you something clear. There's a difference between doing the work and being able to explain it.”

“You're screening for thinking, not code. That's the gap you found, and it's real.”

“The code doesn't match the person walking through it. That gap is exactly the one that matters. When someone hands you polished output and then can't defend it, you're not looking at skill—you're looking at something they borrowed. Might be an AI, might be a colleague they didn't cite, might be Stack Overflow they memorized. Doesn't matter which. The actual work is in explaining it.”

“Her code doesn't match her thinking. That's the red flag, not the nerves.”

“The code is one thing. If she can't explain what she wrote, she didn't write it. I'd give her a chance to walk through it again at a quieter moment, but I wouldn't let that freeze moment vanish from the conversation—ask what happened, ask why it was hard. You're looking at someone who either panicked or copied, and you need to know which one it was before she's sitting in your office on day one.”

“She froze because she doesn't own it. That gap between the code and the explanation is the thing—if the work was hers, walking through it wouldn't be a quiz she fails on. I've done this enough times with nearshore candidates and in-house teams: the ones who wrote the code can tell you why they made each choice, what they'd change now, where they got stuck. The ones who didn't can tell you what it does.”

“You froze her because the code looks too good—but the explanation fell apart. Here's the hard truth: you're testing two different things, and one doesn't tell you much about the other.”