Skip to content
中文Sign in

Recording & AI Use Notice

The short version

  • It only listens after you tap Start on the interview screen, and stops when you end or leave the page.
  • While listening, speech is transcribed in real time and audio is not stored; the transcript and hint cards stay in your round. Interview recordings you upload are deleted by default once transcribed; only the text is kept.
  • To record a real interview, get the interviewer's consent first and respect their company's confidentiality rules.

When it listens

OfferSpar only uses your microphone after you open the interview screen and tap Start. Ending the interview or leaving the page stops it.

With the AI interviewer on (the AI mock interview), it listens only when you answer by voice. With it off, it listens to the interviewer's questions and gives a hint card for each new one.

Outside the interview screen, OfferSpar never accesses your microphone.

How audio is handled

Audio is sent in real time to a speech recognition provider and turned into text. We do not store audio, and it is never written to disk.

We keep the transcript, the questions detected and their hint cards in your round. You can view and delete them.

Text transcribed from the other party's (the interviewer's) speech is used only for your own debrief and is never shared with anyone else [TBC].

For sessions a mentor schedules for you, that mentor can see the record, to coach you.

When others are speaking

If a mentor, peer or anyone else is speaking, tell them before you start listening that the session will be transcribed and kept.

Recording real interviews

After a real interview, you can log the questions and your answers in the round for your debrief. No recording is needed.

If you want to record a real interview, get the interviewer's consent first and respect the company's confidentiality rules. Without consent, don't record, and don't upload recordings with other people's voices to OfferSpar.

You can also upload a recording of a real interview to a round for your debrief. Before uploading you confirm that you have the other party's consent, or that the recording is only for your own debrief.

Uploaded recordings are sent to a transcription provider and turned into text split into the interviewer and you, then a language model pulls out the questions and your answers and adds them to that round's record. Speakers are detected automatically and can be flipped; you can swap them in one click or fix single lines.

The original audio is deleted by default once it's transcribed, and only the text is kept. If transcription fails, it's kept for up to 7 days so you can retry. Only you can see the transcript and the questions pulled from it, and you can edit or delete them any time.

Labeling AI content

Prep, debriefs and hint cards are generated by a language model and labeled in the product. They are practice material and can be wrong, so use your own judgment.

Operator: 天津锐浪科技有限公司 · Contact us support@offerspar.com