CLOUD TRANSCRIPTION

Interview transcription AI: from spoken question to text and answer

InterviewOS Lab transcribes interview audio in real time. Choose Interviewer to capture the other participants through your computer audio, or Me to practise with your microphone. Managed cloud speech-to-text turns each turn into text, the app waits for a complete question, and the transcript feeds the answer workflow in the same session.

Watch a spoken question become a transcript and a grounded answer, then continue with a follow-up in the same session. Uses a fictional example profile.

How it works

How live interview transcription works

  1. Step 1

    Choose what to listen to

    Under Listen to, pick Interviewer for computer audio, which carries the remote participants of a Zoom, Microsoft Teams or Google Meet call, or Me to practise with your own microphone. Then choose the audio device.

  2. Step 2

    Check the signal

    Check your audio listens for six seconds and tells you whether a signal reached the selected device. Nothing is saved or sent for transcription during the check.

  3. Step 3

    Set the language

    Leave Language on Auto-detect or choose the interview language, such as English, Spanish, German, French, Polish, Turkish, Azerbaijani or Japanese, so transcription and answers follow it.

  4. Step 4

    Start listening

    Audio goes to managed cloud speech-to-text, and the transcript appears in the listening bar as the interviewer speaks.

  5. Step 5

    Wait for the whole question

    Question detection waits for a complete question or request; a pause alone does not start an answer. Press Answer now if you want to respond sooner.

  6. Step 6

    Answer from the transcript

    Each detected question goes to the answer workflow with your CV and job description, and the session history is kept on your computer.

In the app

From spoken question
to useful context.

Capture the interviewer through computer audio, or use your microphone for practice. Managed cloud transcription turns the speech into text for the answer workflow.

01

Choose the audio source

Computer audio captures the remote participants. Microphone mode is for self-practice.

02

Read the transcript

Speech is transcribed in the same live session that generates your answers.

03

Understand the processing

The customer app sends audio for managed cloud transcription. Transcription and AI answers require an internet connection.

Audio setup

Capturing computer audio on Windows and macOS

Computer audio is what your speakers or headphones play, which is why it contains the interviewer and not your own microphone.

  • Windows. Computer audio is captured from the output device with no extra driver.
  • macOS 13 and later. System audio is captured natively after you allow Screen & System Audio Recording in System Settings.
  • macOS 11 and 12. A virtual audio driver such as BlackHole is needed, plus a Multi-Output Device so you can still hear the call.
  • Microphone practice. Me mode transcribes your own voice for self-practice; macOS asks for Microphone permission the first time.
Before you start

Requirements and limits

What you need

  • Desktop app. Windows 10 version 2004 or later, or Windows 11; or macOS 11 or later.
  • Internet connection. Transcription and AI answers run on managed cloud services, so the app must be online while you use them.
  • Account and credits. Sign in with an InterviewOS Lab account. 1 credit = 1 minute of live interview, and AI preparation features draw on the same balance. Package credits are on the pricing page.
  • macOS permissions. Screen & System Audio Recording for computer audio on macOS 13 and later, and Microphone for practice mode.

Limits to know

  • No offline transcription. The customer app sends audio to managed cloud speech-to-text, so it needs an internet connection.
  • Audio quality matters. Crosstalk, background noise and heavy compression reduce accuracy, so check names and numbers.
  • Computer audio means all of it. Everything your computer plays is captured, so mute other apps during a session.
  • Noise is filtered. Unclear fragments are held back rather than answered, which can occasionally drop a very short turn.
  • Follow the interview rules. Use live assistance only where the interview or assessment rules allow it, and get any consent needed for recording or transcription.
FAQ

Transcription questions.

Short answers about what the app does and does not do.

Does interview transcription work offline?

No. The customer app sends audio to managed cloud speech-to-text, so transcription and AI answers need an internet connection.

Whose voice does it transcribe?

The source you choose. Interviewer captures your computer's audio output, which carries the other participants on the call; your own microphone is not part of that capture. Me captures your microphone for practice.

Do I need extra software on a Mac?

On macOS 13 or later, no: system audio is captured natively once you allow Screen & System Audio Recording. On macOS 11 and 12 you need a virtual audio driver such as BlackHole and a Multi-Output Device.

Which languages can it transcribe?

Language can be left on Auto-detect or set to a specific interview language, including English, Spanish, German, French, Portuguese, Italian, Dutch, Russian, Ukrainian, Polish, Turkish, Azerbaijani, Hindi, Chinese, Japanese, Korean and Arabic.

How is transcription billed?

Transcription is part of a live session, which uses credits at 1 credit = 1 minute of live interview. The credits in each package are listed on the pricing page.

Get started with InterviewOS Lab

Your next interview.
A clearer way through.

Try it in a real interview: 60 free credits, up to 60 minutes of live interview. No credit card required.