Explore Further Issues

The "Explore_Further_Issues" tracker detects moments in a conversation where an agent asks if there are any other issues, questions, or concerns the participant wants to raise. Recognizing these proactive inquiries highlights how representatives go beyond the immediate problem, ensuring comprehensive support and enhancing overall customer experience.

Who Benefits:

• Project Managers & Team Leads: Understand how often and effectively agents probe for additional concerns, guiding coaching on thoroughness.
• Sales & CX Teams: Confirm that agents don’t just solve the first issue but also ensure no hidden problems remain, improving satisfaction and retention.
• Business Managers & Strategists: Spot patterns in follow-up inquiries, helping refine processes, scripts, and training materials to meet evolving needs.
• Compliance Officers: Ensure that agents follow protocols encouraging comprehensive assistance, maintaining service quality and trust.

Value Proposition:

• Enhanced Customer Care: By detecting when agents explore further issues, teams can measure how well they ensure customers leave with all concerns addressed.
• Continuous Improvement: Insights from these interactions feed into coaching, helping representatives consistently deliver top-tier support.
• Actionable Data: Structured JSON output seamlessly integrates into dashboards, enabling trend analysis and informed decision-making.

Data Dictionary

Objective:

Detect phrases in a conversation indicating the agent is proactively checking if there are additional issues or concerns to address. The tracker returns occurrences of matched phrases, along with timestamps and confidence.

Proposed JSON Schema:

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

Data Dictionary:

If No Matches Found:
If no such inquiries are detected, return an empty matches array for the "Explore_Further_Issues" tracker or omit the tracker. If no trackers match, return:

{
  "trackers": []
}

Determining Confidence & Thresholds

How the Explore_Further_Issues Tracker Works:

• Seed the tracker with phrases representing a check for additional problems (e.g., "Anything else bothering you?", "Any other issues to address?", "Is there anything else I can assist with?").
• The system interprets these phrases and detects similar statements.
• Confidence reflects how closely the detected phrase aligns with asking about further issues.

Confidence & Calibration:

• Initial Testing: Validate on known calls to ensure that detected phrases accurately represent inquiries for more issues.
• Refine Phrases: Add or remove seed phrases to improve coverage and accuracy.
• Thresholding: Downstream consumers can filter out low-confidence matches as needed.
• Stakeholder Feedback: Involve CX leads or compliance officers to refine definitions and ensure relevance.

Prompt Construction & Instructions

Role Specification & Reiteration:

• Present the system as a “highly experienced assistant” identifying when an agent checks for additional concerns.
• Reiterate instructions: return only JSON, follow the schema, no extra commentary.

No Hallucination:

• Only detect relevant phrases if present.
• If uncertain, assign lower confidence or no match.

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 'Explore_Further_Issues' mentions in a conversation. You have:

• A transcript
• A tracker named 'Explore_Further_Issues' defined by seed phrases indicating checking for additional problems.

Your task:

1. Detect phrases where the agent asks if there are more issues/concerns.
2. Return text, timestamp, confidence.
3. Return only JSON as per schema, no commentary.

Instructions & Variables (example):

{
  "tracker": {
    "name": "Explore_Further_Issues",
    "phrases": [
      "anything else I can help you with",
      "any other issues",
      "any other concerns",
      "anything else bothering you"
    ]
  }
}

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

[TRANSCRIPT_JSON][TRANSCRIPT_JSON]"

Expected Output Format:

{
  "trackers": [
    {
      "name": "Explore_Further_Issues",
      "matches": [
        {
          "text": "Is there anything else I can help you with today?",
          "timestamp": "2024-07-10T10:15:00Z",
          "confidence": 0.9
        }
      ]
    }
  ]
}