AI for interviews

The recruiter asks you to walk through a time you handled conflict on a team. Your mind goes to three half-stories at once and none of them are landing. Whisply hears the question through your Mac microphone, drops a tight STAR-format suggestion in the overlay, and you read one phrase, look back at the camera, and tell the story in your own voice. The pause that would have been a long stare becomes a normal beat of thinking.

The work happens in a small panel you summon with Cmd+Return. It is not a teleprompter. The phrasing is yours, the judgment is yours, and the assistant only fills in the thing you almost remembered. For a system design round it might surface the trade-off you forgot between read-heavy and write-heavy workloads. For a fit interview it might remind you which of the company values maps to the story you started telling. The help is small and well-timed, which is the only kind that works under pressure.

Because Whisply runs on your machine and not as a bot in the call, the interviewer never sees a second participant. There is no logo on the recording, no banner in the participant list, no Notetaker icon sitting in the corner getting on everyone's nerves. From the other side of the call you look like a candidate who prepared.

Behavioral rounds are pattern matching on stories. Whisply listens for cues like tell me about a time, walk me through, describe a situation, and surfaces a STAR scaffold pulled from the notes and resume you loaded earlier. You glance, you remember which project fits, you tell it like a person and not like a script. The overlay disappears the second you start talking confidently.

Technical rounds are different. For a coding screen you can ask Whisply for the time complexity of the approach you just sketched, or for a sanity check on an edge case before you write it. For a system design conversation it can hold the diagram in mind while you think out loud, prompting you with the next layer to consider. None of it shows up in the shared screen, so the interviewer sees your editor and your face, not your help.

Culture-fit questions reward specificity. Whisply pulls from the company research you pasted in before the call, so when the question is why us, your answer cites the actual product launch from last quarter instead of a generic compliment. The signal you give off is that you did the homework, because you did, and the assistant just made sure the right detail was on the tip of your tongue.

The overlay is marked with macOS content protection, which means Zoom, Google Meet, Microsoft Teams, Webex, and most browser-based call tools will not capture it in a screen share or a cloud recording. If the recruiter is sharing their screen to walk you through a coding problem, your hints stay on your side. If you are sharing your screen to demo a project, the panel does not appear in the stream the interviewer sees.

On a second monitor or with the panel tucked off to the side, the assist sits in your eyeline without dragging your eyes away from the camera. Pro Undetected adds Computer Use through Accessibility permission, so for take-home style live exercises the assistant can act on your Mac when you ask it to, not autonomously and not behind your back.

The privacy story matters because interviews are personal. Your conversations are not piped to a third party to be stored and mined. The recruiter is not being recorded by a bot you let into the room. The help is for you, the discretion is for everyone.

Generic AI advice in an interview is worse than no AI advice. Whisply gets useful when you give it your resume, the job description, two or three stories you want to land, and a few notes on the company. Paste it into the overlay before the call. From that point on the suggestions are grounded in your actual experience instead of a stock answer that any candidate could read.

For technical screens, drop in the role's stack and the kinds of systems the team works on. For fit interviews, paste a recent blog post from the company or a quote from the hiring manager you found on LinkedIn. The overlay treats this as the working set for the session. When the interviewer asks why this role, the answer that surfaces is specific to the team and the work, not a polished nothing.

After the call, Whisply can produce a debrief with the questions it heard, the moments you handled well, and the ones where the suggestion landed late. The debrief is for you, not for the recruiter, and it gets sharper for the next round.