The obvious version of Mokchi is voice-first.
You get asked a question. You speak. The interviewer listens and responds.
That’s the closest thing to the actual environment we’re trying to prepare for.
But I kept wanting to use the product in situations where talking out loud made no sense.
Sometimes I’m working. Sometimes I’m thinking through a question casually. Sometimes I just want to test whether the interviewer understands the substance of an answer.
So we added text.
The important decision was not to treat it like a second-class fallback.
Voice and text should both go through the same interviewer, evaluator, follow-up logic, and debrief. The input changes. The intelligence shouldn’t.
Voice helps rehearse the actual performance.
Text helps work through the thinking.
I think both are useful, just for slightly different moments.