Introduction - Voice diarization

Voice diarization in Qik Office automatically identifies, separates, and labels individual speakers within a meeting. It ensures that every spoken statement is attributed to the correct participant, creating a clear, structured, and speaker-wise record of the entire conversation.

This capability allows Qik Office to transform meeting data into organized, meaningful, and traceable information. By accurately linking each statement to its speaker, it removes ambiguity in meeting documentation and enables reliable transcripts, precise meeting minutes, and improved accountability across teams.

Purpose

The primary purpose of voice diarization is to convert raw meeting audio into structured, speaker-identified information. This allows Qik Office to transform conversations into usable outputs such as:

• Speaker-labeled transcripts
• Automated meeting minutes
• Task assignments
• Meeting insights and analysis

How voice diarization works?

1. Voice sample registration Each user records a voice sample by reading a provided text. This sample is used to create the user's unique voice profile.
2. Voice profile creation The system analyzes the sample and stores a voice signature associated with the user’s account.
3. Meeting audio analysis When a meeting is recorded, the system processes the audio and detects different speech segments.
4. Speaker identification The detected speech is matched with stored voice profiles, and the transcript is labeled with the correct participant names.

Why voice diarization in in-person meetings?

1. Makes meeting minutes context-aware, not just text-based
2. Enables actionable insights, such as:
   • Who raised which point
   • Speaker-wise summaries
   • Cleaner follow-ups and accountability

How to give voice sample ?

1. Click the "Add voice sample" icon in the top header to begin your personalized setup.

2. Click enable microphone to grant the app necessary permissions and click "Get started".

3. Tap the "Record" button and speak the phrase clearly to capture your voice.

4. Read the required consent phrase on the screen to ensure your voice recorded.

5. Wait for the system to validate your sample by analyzing the audio and finalizing your voice profile.

6. Confirm the voice by clicking "Done" once the green checkmark appears.

7. A green waveform badge on your display picture to confirming that voice identification is live and active.