AI
Interview Docs

Transcripts & Analysis

Interview transcripts are the core output of your campaigns. Learn how to access, analyze, and export transcripts effectively.

Accessing Transcripts

Individual Transcript

View a single interview transcript:

  1. Navigate to Sessions
  2. Filter by campaign or date range
  3. Click on a completed session
  4. View full transcript with timestamps

Bulk Access

Access multiple transcripts:

  • Campaign View: See all transcripts from a campaign
  • Project View: All transcripts across multiple campaigns
  • Search: Find transcripts containing specific keywords
  • API: Programmatic access via REST API

Transcript Format

Structure

Each transcript contains:

Header Information:

  • Session ID and timestamp
  • Campaign and script name
  • Participant email (if not redacted)
  • Interview duration
  • Completion status

Conversation:

[00:00:12] AI: Hello! Thank you for joining today. How are you doing?
[00:00:15] Participant: I'm doing well, thanks for asking.
[00:00:18] AI: Great! Let's get started. Can you tell me about...

Footer:

  • Total duration
  • Word count
  • Turn count
  • Sentiment summary

Turn-by-Turn Data

Each turn includes:

  • Timestamp: When the turn started
  • Speaker: AI or Participant
  • Text: What was said
  • Confidence: Speech recognition confidence (0-100%)
  • Sentiment: Positive, neutral, or negative
  • Duration: How long the turn lasted

Viewing Options

Standard View

Clean, readable format for quick review:

  • Formatted like a conversation
  • Color-coded speakers
  • Inline timestamps
  • Click timestamp to play audio at that point

Detailed View

Additional metadata visible:

  • Confidence scores
  • Sentiment indicators
  • Audio waveforms
  • Turn durations
  • AI decision points (why certain follow-ups were asked)

Side-by-Side Audio

Synchronized audio playback:

  • Transcript auto-scrolls with audio
  • Click any line to jump to that point
  • Playback speed control (0.5x - 2x)
  • Highlight current speaker

Analysis Features

Automatic Insights

AI-generated analysis:

Summary:

  • Key points discussed
  • Main topics covered
  • Notable quotes
  • Overall sentiment

Topics Extracted:

  • Automatically identified themes
  • Frequency of mentions
  • Sentiment per topic

Action Items:

  • Tasks or follow-ups mentioned
  • Commitments made
  • Questions left unanswered

Automatic insights use GPT-4 to analyze transcripts. Available on Pro and Enterprise plans.

Sentiment Analysis

Track emotional tone throughout the interview:

  • Overall Sentiment: Positive, neutral, or negative
  • Sentiment Over Time: Graph showing emotional progression
  • Turning Points: Where sentiment shifted dramatically
  • Engagement Level: How enthusiastic participant responses were

Keyword Highlighting

Find specific terms across transcripts:

  1. Enter keywords or phrases
  2. See all mentions highlighted
  3. View context around each mention
  4. Export matching excerpts

Use cases:

  • Track product mentions
  • Find pain points
  • Identify competitors mentioned
  • Discover feature requests

Exporting Transcripts

PDF Export

Professional, shareable format:

  • Formatted header with session metadata
  • Clean conversation layout
  • Optional: Include analysis summary
  • Optional: Redact participant information

Best for:

  • Sharing with stakeholders
  • Including in reports
  • Archival purposes

CSV Export

Spreadsheet-compatible format:

timestamp,speaker,text,sentiment,confidence
00:00:12,AI,"Hello! Thank you for joining today.",neutral,98
00:00:15,Participant,"I'm doing well thanks.",positive,95

Best for:

  • Quantitative analysis
  • Importing into analytics tools
  • Custom processing

JSON Export

Programmatic access to all data:

{
  "session_id": "ses_123",
  "started_at": "2024-10-20T10:00:00Z",
  "turns": [
    {
      "timestamp": 12.5,
      "speaker": "ai",
      "text": "Hello! Thank you for joining today.",
      "sentiment": "neutral",
      "confidence": 0.98
    }
  ]
}

Best for:

  • API integrations
  • Custom analysis pipelines
  • Data science workflows

Audio Export

Download interview audio:

  • MP3: Compressed, smaller file size
  • WAV: Uncompressed, high quality
  • Separate Tracks: AI and participant on different channels

Search and Filtering

Find transcripts by:

  • Keyword: Search transcript content
  • Participant: Find specific participant's interviews
  • Date Range: Filter by completion date
  • Campaign: View all from a campaign
  • Status: Completed, abandoned, or in-progress

Advanced Filters

Combine multiple criteria:

Campaign: "Q4 Customer Feedback"
AND Date: Last 30 days
AND Contains: "pricing" OR "expensive"
AND Sentiment: Negative

Saved Searches

Save frequently used queries:

  • One-click access
  • Share with team members
  • Auto-execute on schedule
  • Email results

Collaborative Analysis

Comments and Tags

Add context to transcripts:

Comments:

  • Highlight specific turns
  • Add notes and interpretations
  • Tag team members for review
  • Thread discussions

Tags:

  • Apply custom labels
  • Filter by tags
  • Track common themes
  • Organize for reporting

Sharing

Share transcripts securely:

  • Internal: Share with team members
  • External: Generate time-limited share links
  • Permissions: View-only or allow commenting
  • Redaction: Auto-redact PII before sharing

Aggregate Analysis

Cross-Transcript Insights

Analyze multiple interviews together:

Common Themes:

  • Topics mentioned across interviews
  • Frequency and importance
  • Sentiment per theme

Quote Collections:

  • Group related quotes
  • Find representative examples
  • Export for presentations

Comparative Analysis:

  • Compare segments (e.g., happy customers vs. churned)
  • A/B test results across script variations
  • Trend analysis over time

Reporting

Generate reports from multiple transcripts:

Executive Summary:

  • Key findings
  • Participant demographics
  • Top quotes
  • Recommendations

Detailed Report:

  • Full analysis methodology
  • All transcripts included
  • Statistical breakdowns
  • Appendix with raw data

Custom Reports:

  • Build your own template
  • Choose metrics to include
  • Schedule automatic generation

Pro Tip

Tag transcripts as you review them (e.g., "feature request", "bug report", "positive feedback"). This makes it easy to filter and generate focused reports later.

Privacy and Redaction

PII Detection

Automatically detect personally identifiable information:

  • Email addresses
  • Phone numbers
  • Social Security Numbers
  • Credit card numbers
  • Physical addresses
  • Names (optional)

Redaction Options

Choose how to handle detected PII:

Mask: Replace with placeholders

Email: [EMAIL_REDACTED]
Phone: [PHONE_REDACTED]

Remove: Delete entirely

My email is  and you can call me at .

Encrypt: Store encrypted, decrypt on-demand with permission

Manual Redaction

Override automatic detection:

  1. Select text in transcript
  2. Choose Redact
  3. Select redaction type
  4. Apply to all exports

API Access

REST API Endpoints

Programmatic transcript access:

GET /api/sessions/{id}/transcript
GET /api/campaigns/{id}/transcripts
POST /api/transcripts/search

See API Reference for details.

Webhooks

Real-time notifications:

  • transcript.completed: When transcript is ready
  • transcript.analyzed: When AI analysis completes
  • transcript.exported: When export is generated

Configure webhooks in Settings → Webhooks.

Best Practices

Review Process

Establish a workflow:

  1. Initial Review: Skim all transcripts for quality
  2. Tagging: Apply relevant tags
  3. Deep Analysis: Detailed review of key transcripts
  4. Quote Extraction: Pull notable quotes
  5. Reporting: Synthesize findings

Quality Assurance

Check transcript accuracy:

  • Spot Check: Review 10% of transcripts with audio
  • Confidence Scores: Flag low-confidence turns for review
  • Nonsensical Responses: May indicate speech recognition errors
  • Report Issues: Help improve our models

Organization

Keep transcripts organized:

  • Naming Conventions: Use descriptive campaign names
  • Folder Structure: Organize by project/timeframe
  • Archive: Move old transcripts to archive
  • Retention Policy: Delete after compliance period expires

Troubleshooting

Transcript Missing or Incomplete

Possible causes:

  • Session abandoned mid-interview
  • Technical error during recording
  • Audio quality too poor to transcribe

Solutions:

  • Check session status (may still be processing)
  • Review audio file (may be partial recording)
  • Contact support for recovery attempt

Inaccurate Transcription

Common issues:

  • Heavy accent or dialect
  • Background noise
  • Multiple speakers
  • Technical jargon

Solutions:

  • Review with audio for context
  • Manually correct important sections
  • Report consistently problematic words to improve our models

Export Failed

Troubleshooting:

  • Check file size limits
  • Verify export format support
  • Try smaller batch size
  • Check browser console for errors
Previous
Interview Sessions
Next
Billing & Credits