Skip to content

Transcription

Dialer.CX can automatically transcribe call recordings into searchable text using AI-powered speech recognition. Transcripts are linked to recordings, searchable across the platform, and can power downstream analysis features like sentiment scoring and compliance checking.

AI Settings — Transcription tab configuration

  1. A call is recorded during the conversation
  2. The recording is sent to your configured transcription provider
  3. The AI generates a text transcript with speaker identification
  4. The transcript is stored and linked to the recording
  5. Downstream features (analysis, summaries, scoring) can run on the transcript

The transcript identifies who is speaking throughout the call:

  • Agent speech labeled separately from customer speech
  • Easy to scan conversations and find specific parts
  • Color-coded speaker labels in the playback view

In the Recording Center, transcripts are tightly coupled with audio playback:

  • Transcript highlights in real time as the recording plays
  • Click any line in the transcript to jump to that moment in the audio
  • Makes quality review and coaching significantly faster

Full-text search transforms how you find specific calls:

  • Search across all transcripts from the Recording Center
  • Find calls where specific topics, products, or phrases were discussed
  • Use search to build coaching playlists or compliance audit sets

Recording Center — transcript view with synchronized audio playback

Navigate to Admin > AI Settings > Transcription:

Sidebar navigation to AI Settings

SettingDescription
ProviderSelect your transcription provider
ModelChoose model variant (standard vs. enhanced accuracy)
LanguagePrimary language for recognition
Auto-TranscribeAutomatically transcribe all recordings, or on-demand only
OptionSpeedBest For
StandardNormalHigh accuracy, broad language support
FastAcceleratedHigh throughput, lower latency

If auto-transcription is off, you can transcribe individual recordings:

  1. Open a recording in the Recording Center
  2. Click Transcribe
  3. Wait for processing (typically 30-60 seconds)
  4. Transcript appears alongside the audio player

Recording Center — full page with transcript and analysis panels