gong_get_transcript

Retrieve detailed transcripts from Gong calls with speaker attribution
READ

gong_get_transcript

Retrieves the complete transcript of a Gong call with full speaker attribution. This is the primary tool for extracting conversation data to create minds or analyze sales conversations.

Gong transcripts include accurate speaker attribution with both names and email addresses, making them ideal for creating minds and analyzing individual communication patterns.

Prerequisites

Gong Integration Required

This tool requires Gong integration to be configured. See the Gong Setup Guide for configuration instructions.

Parameters

callId

Typestring
RequiredYes
DescriptionThe unique call identifier from gong_list_calls response
Examplecall_abc123

Request

1{
2 "callId": "call_abc123"
3}

Response

1{
2 "callId": "call_abc123",
3 "title": "Discovery Call - Enterprise Prospect",
4 "date": "2024-11-15T14:30:00Z",
5 "duration": 3600,
6 "transcript": [
7 {
8 "speakerName": "Mike Chen",
9 "speakerEmail": "mike.chen@yourcompany.com",
10 "text": "Thanks for taking the time today. I'd love to learn more about your current challenges with customer onboarding.",
11 "timestamp": "00:00:15"
12 },
13 {
14 "speakerName": "Sarah Johnson",
15 "speakerEmail": "sarah.johnson@prospect.com",
16 "text": "Absolutely. Right now, our biggest pain point is that it takes 3-4 weeks for new customers to see value. We're losing about 20% of customers during onboarding.",
17 "timestamp": "00:00:28"
18 },
19 {
20 "speakerName": "Mike Chen",
21 "speakerEmail": "mike.chen@yourcompany.com",
22 "text": "That's significant. Can you walk me through your current onboarding process?",
23 "timestamp": "00:00:45"
24 }
25 ]
26}

Response Properties

PropertyTypeDescription
callIdstringThe call identifier
titlestringCall title from Gong
datestringISO 8601 timestamp of the call
durationnumberCall duration in seconds
transcriptarrayArray of transcript segments in chronological order
transcript[].speakerNamestringFull name of the speaker
transcript[].speakerEmailstringEmail address of the speaker
transcript[].textstringThe spoken text for this segment
transcript[].timestampstringTimestamp within the call (HH:MM:SS format)

Common Use Cases

Extract conversation data to train digital minds on actual sales interactions.

Example:

"Get the transcript from my Gong call with Enterprise Corp
and create a mind for the prospect"

Natural language workflow:

  1. AI uses gong_list_calls to find the call
  2. Uses gong_get_transcript to retrieve full conversation
  3. Creates mind with create_mind
  4. Uploads transcript as training data
  5. Creates snapshot and trains the model

What you get:

  • Mind trained on actual prospect communication style
  • Ability to simulate future conversations
  • Insights into objection handling and decision-making patterns

Extract individual speaker segments to understand communication styles.

Example:

"Get the transcript from call_abc123 and analyze how
the prospect responded to pricing discussions"

Natural language workflow:

  • AI retrieves full transcript
  • Filters to prospect speaker segments
  • Analyzes segments containing “pricing” keywords
  • Identifies patterns and sentiment

What you get:

  • Speaker-specific conversation segments
  • Communication style insights
  • Objection patterns
  • Decision-making indicators

Use real call transcripts to train sales teams on effective conversations.

Example:

"Get transcripts from my top 5 performing calls this quarter
and extract the discovery question patterns"

Natural language workflow:

  • AI lists calls and identifies top performers
  • Retrieves transcripts for each call
  • Analyzes question patterns
  • Extracts effective discovery techniques

What you get:

  • Real-world examples of effective techniques
  • Question frameworks that work
  • Objection handling approaches
  • Closing strategies

Error Responses

1{
2 "error": "Not Found",
3 "message": "Call with ID 'call_abc123' does not exist"
4}

Solution: Verify the call ID is correct. Use gong_list_calls to get valid call IDs from your account.

1{
2 "error": "Unauthorized",
3 "message": "Gong authentication failed"
4}

Solution: Verify your Gong credentials are configured correctly. See the Gong Setup Guide.

1{
2 "error": "Forbidden",
3 "message": "You do not have permission to access this call"
4}

Solution: Check your Gong user permissions. You may not have access to this specific call or recording.

1{
2 "error": "Bad Request",
3 "message": "Transcript not yet available for this call"
4}

Solution: The call may still be processing. Wait a few minutes and try again. Check the call status in Gong’s web interface.

Best Practices

Speaker Filtering
  • Extract segments for specific speakers when creating individual minds
  • Use speaker emails to accurately identify participants
  • Filter internal vs. external speakers based on email domains
  • Create separate minds for each unique speaker
  • Transcripts can be large—consider filtering before uploading to Mind Reasoner
  • Remove filler words and repetitions for cleaner training data
  • Focus on substantive conversation segments
  • Maintain speaker attribution when processing
  • Use transcript text directly as training data for minds
  • Combine multiple call transcripts for richer training
  • Filter to specific topics or conversation types
  • Preserve speaker context (name, email, role)

Integration with Mind Reasoner

Creating a Mind from a Gong Transcript

1

List Available Calls

Use gong_list_calls to find calls with the person you want to create a mind for

2

Get Transcript

Use gong_get_transcript to retrieve the full conversation

3

Extract Speaker Segments

Filter transcript to segments from the target speaker

4

Create Mind

Use Mind Reasoner’s create_mind with the speaker’s details

5

Upload Transcript

Upload the filtered transcript as training data

6

Train

Create snapshot and wait for training to complete

Natural language example:

"Create a mind for Sarah Johnson using all my Gong calls
with her from the last quarter"

AI handles the entire workflow automatically:

  • Lists calls with Sarah
  • Retrieves all transcripts
  • Extracts Sarah’s segments
  • Creates mind and trains on her communication patterns

See Also