Introductions

The "Introductions" tracker detects phrases that indicate a participant is introducing themselves, another person, or providing relevant identifying information. By capturing these moments, teams gain insights into call beginnings, relationship building, and whether participants follow scripts or policies for proper introductions.

Who Benefits:

• Project Managers & Team Leads: Understand if representatives consistently start calls by stating their name and role, improving professionalism and trust.
• Sales & CX Teams: Recognize how well agents introduce themselves, impacting first impressions and establishing credibility from the outset.
• Business Managers & Strategists: Spot patterns in how introductions occur, informing refinements to onboarding scripts, training materials, or brand positioning.
• Compliance Officers: Ensure required identity disclosures, roles, or disclaimers are mentioned at the start of interactions.

Value Proposition:

• Clear Beginnings: Highlighting introductions ensures that calls start on a positive, transparent note, fostering trust and understanding.
• Scalable Insights: Structured data integrates easily with dashboards, aiding analysis of introductory compliance, best practices, and variations across teams.
• Continuous Improvement: Understanding how introductions are made guides iterative enhancements to scripts, training modules, and communication guidelines.

Data Dictionary

Objective:

Detect phrases in the conversation that suggest a participant is introducing themselves or someone else, stating a name, role, or affiliation. The tracker returns occurrences of matched phrases, along with timestamps and confidence. This data helps teams confirm that calls start with proper identification and adherence to any introduction protocols.

Proposed JSON Schema:

{
  "trackers": [
    {
      "name": "Introductions",
      "matches": [
        {
          "text": "<string>",
          "timestamp": "<ISO 8601 or relative time>",
          "confidence": <number> // 0.0 to 1.0
        }
      ]
    }
  ]
}

Data Dictionary:

If No Matches Found:
If no introductions are detected, return an empty matches array for "Introductions" or omit the tracker. If no trackers match at all, return:

{
  "trackers": []
}

Determining Confidence & Thresholds

How the Introductions Tracker Works:

• Seed the tracker with phrases commonly used to introduce oneself or another party (e.g., "my name is," "I’m [Name] from [Department/Company]," "let me introduce").
• The system generalizes these phrases to identify semantically related statements indicating an introduction.
• Confidence indicates how strongly the detected phrase aligns with an introduction scenario.

Confidence & Calibration:

• Initial Testing: Validate on known calls to ensure detected phrases accurately reflect actual introductions, not merely mentioning names without self-identification.
• Refine Phrases: Add or remove seed phrases to capture cultural or organizational variations in introductions.
• Thresholding: Downstream consumers can filter out low-confidence matches if needed.
• Stakeholder Feedback: Incorporate insights from CX leads or compliance officers to refine what qualifies as a valid introduction.

Prompt Construction & Instructions

Role Specification & Reiteration:

• Present the system as a “highly experienced assistant” identifying introduction moments.
• Reiterate instructions: return only JSON, no extra commentary, follow schema strictly.

No Hallucination:

• Only detect introductions if the transcript provides clear evidence of self-identification or formal introduction.
• If uncertain, assign lower confidence or omit matches.

Strict Formatting:

• Return only JSON with trackers.
• If none found, return empty trackers.

Example Prompt for Implementation

System Message (Role: System):
"You are a highly experienced assistant that identifies 'Introductions' mentions in a conversation. You have:

• A transcript
• An 'Introductions' tracker defined by seed phrases indicating self or group introductions.

Your task:

1. Detect phrases where a participant states their name, role, or affiliation at the start or when joining a conversation.
2. Return text, timestamp, and confidence.
3. Return only JSON per the schema, no extra commentary.

Instructions & Variables (example):

{
  "tracker": {
    "name": "Introductions",
    "phrases": [
      "my name is",
      "I’m [Name]",
      "let me introduce",
      "I’m calling from",
      "I’m with [Company]"
    ]
  }
}

User Message (Role: User):
"Analyze the following transcript and detect 'Introductions' mentions:

[TRANSCRIPT_JSON][TRANSCRIPT_JSON]"

Expected Output Format:

{
  "trackers": [
    {
      "name": "Introductions",
      "matches": [
        {
          "text": "Hi there, my name is Alex and I’m from the marketing team.",
          "timestamp": "2024-07-10T09:05:00Z",
          "confidence": 0.9
        }
      ]
    }
  ]
}