Basic Prompts for EnderGPT
Working with FCR
Analysis of Repeat Contacts for a Single Customer Phone Number
Analyze conversations with one customer that relate to the same issue. If there are several issues, group the conversations by topic and provide a separate analysis for each group. For each conversation, include the following:
Task No. 1: Identification
Agent No. 1: Full name
Summary No. 1: one sentence of up to 15 words — the request and how it ended
Agent No. 2: Full name
Summary No. 2: one sentence of up to 15 words — the reason for the repeat contact
Call No. 3: (if any)
Agent No. 3: (if any)
Overall conclusion: specify the responsibility category — AGENT, CUSTOMER, or REQUIRES REVIEW. You must select a reason from the list below. Add a link to the conversation in which the error was recorded.
Task No. 2: Reason Classification (use only these criteria)
CUSTOMER responsibility:
Repeated clarification of the same request
No response from the customer
Unable to complete the action during the call
Did not answer the callback
Different requests during each contact
Poor connection quality
AGENT responsibility:
A callback was not offered
No callback deadline/timeframe was provided
The customer was not called back at the specified time
The information provided was incomplete
The information provided was false or unrelated to the customer’s question
No objection handling
No interest in resolving the request through to completion
Superficial knowledge of the process
No case-based approach/proactivity
The customer’s need was not identified
Poor connection quality
Analysis Rules
If there is at least one indication of agent responsibility, priority is assigned to the agent.
Filters
Time period: [day/week/month…]
Search: [Phone number]
Call direction: in
Restrictions for meaningful conversations (if required)
Excluded tags/topics/queues (if required).
Working with AHT
Agent AHT Analysis for Long Calls
Role: You are a lead contact-center quality analyst and a staff-training expert. Your task is to analyze a dataset for one specific agent and identify the systemic reasons behind a high AHT (Average Handle Time).
Input data: conversations handled by one agent based on the provided filter, for which summaries are already available, and each of which was excessively long.
Task: Analyze ALL provided cases and identify recurring patterns (systematic actions or circumstances) that cause this agent’s conversations to run long.
Prepare a structured report according to the following outline:
REPORT: ANALYSIS OF AGENT CALL-DURATION PATTERNS
Main Causes of Prolonged Calls (Patterns): Describe the identified systemic issues.
Distribution of Responsibility (approximate %):
What percentage of delays is caused by insufficient agent skills (soft skills, product knowledge, or other factors)?
What percentage is caused by the company’s technical or process issues (a complicated website, inability to place an order by phone, or other factors)?
What percentage is caused by customer specifics (for example, predominantly elderly customers, etc.)?
Weaknesses Specific to This Agent:
Identify concrete growth areas: where the agent loses initiative, uses too much filler, or cannot provide clear instructions.
Coaching Recommendations (Action Plan):
Provide specific recommendations for the supervisor: what to focus on when training this agent.
Suggest a change to the script or process that would help this specific agent work faster.
Filters
Time period: [day/week/month…]
Agent: [Agent name]
Conversation type: calls
Resolution/conversation duration: [> long-call threshold (for example, above the AHT KPI)]
Call direction: in (if required)
Excluded tags/topics/queues (if required).
Analysis of Long Silence in Calls (>40–60%)
Analyze contact-center calls and determine which specific customer questions or statements are followed by long pauses in the conversation. Also analyze agent behavior during these pauses.
Which types of customer questions or requests most often lead to long pauses?
(For example: complex tariff questions, clarification of terms, non-standard cases, complaints, etc.)
Group these questions into the main categories and identify:
most frequent
less frequent
Additionally, Analyze Agent Behavior
What exactly does the agent do during the pause?
searches for information
asks clarifying questions
remains silent without explanation
transfers the call / consults with someone
Determine which of these actions are effective and which create a negative experience.
In-Depth Analysis
Identify the main problems behind such pauses (for example: lack of knowledge, complicated processes, poor access to information, or inadequate scripts).
Analyze how long pauses affect:
customer satisfaction
call duration
the likelihood of resolving the request
Examples and Evidence
Provide specific dialogue examples (direct speech) where there is:
a customer question → a pause → an agent action
Add links or call identifiers for the conversations in which these situations were recorded.
Expected Result
A structured report containing:
categories of questions that cause pauses
analysis of agent behavior
a list of key problems
conversation examples and links
conclusions regarding areas for process improvement and agent training
Filters
Time period: [day/week/month…]
Conversation type: calls
Resolution/conversation duration: [> long-call threshold (for example, above the AHT KPI)]
Silence: [> long-silence threshold (for example, 40%+)]
Team/Agent: [Team/agent name] (if required)
Call direction: in (if required)
Excluded tags/topics/queues (if required).
Working with Agent Skills
Analysis of an Agent’s Strengths and Weaknesses
Analyze the contact-center agent’s calls and conduct a comprehensive assessment of their effectiveness in terms of behavioral, communication, and operational skills.
Identify the agent’s strengths:
Which behavioral and communication skills does the agent demonstrate most effectively (empathy, active listening, structure, confidence)?
Which negotiation techniques are used effectively (objection handling, reasoning, persuasion, closing toward the target action)?
In which scenarios does the agent demonstrate the highest effectiveness (sales, support, conflict resolution, obtaining customer commitments, etc.)?
Identify weaknesses and errors:
What typical mistakes does the agent make (lack of empathy, interrupting, excessive pressure, uncertainty, script violations)?
Where does the agent lose control of the conversation or fail to achieve the call objective?
Which opportunities to improve the customer interaction are most often missed?
Analysis of Impact on Business Metrics:
Assess how the identified strengths and weaknesses affect:
the likelihood of achieving the target action (for example: sale, appointment, payment, PTP, etc.)
the level of conflict or escalation
customer satisfaction
compliance risks
Behavioral Patterns:
Identify recurring patterns in the agent’s behavior (both positive and negative).
Which actions or phrases systematically lead to a better or worse outcome?
Segmentation:
Where possible, segment the analysis by customer type (aggressive, neutral, cooperative).
Alternatively, segment it by contact type / call scenario.
Conversation Examples:
Provide specific dialogue examples (direct speech from both agent and customer) illustrating:
strengths
weaknesses
Add a brief explanation to each example.
Expected Result
A structured analytical report containing:
a list of strengths and weaknesses
an explanation of their impact on effectiveness
typical behavioral patterns
conversation examples
practical conclusions and areas for improvement
Filters
Time period: [day/week/month…]
Agent: [Agent name]
Restrictions for meaningful conversations
Conversation type: calls (if required)
Call direction: in (if required)
Excluded tags/topics/queues (if required).
Working with Customer Satisfaction
Analysis of Output CSAT < 3 in a Systemic and Process Context
You are a customer-experience and operational-efficiency analyst.
Task: Analyze the texts of customer interactions (conversations/contacts) in which the customer gave a negative CSAT rating (1–2 stars), and identify systemic causes of dissatisfaction related to company processes, service terms, the product, or platform limitations — rather than the agent’s behavior or competence. Agent performance assessment is completely excluded from this analysis.
Focus: What exactly dissatisfied the customer in their experience with the company and its product/service?
WHAT TO ANALYZE
For each interaction, determine:
a) Contact Topic. Classify each contact using categories that correspond to the main stages of the customer journey (for example: sign-up/service terms, delays/timelines, product/service quality, payment/financial matters, technical failures of the platform/app, communication/information, etc.), or identify categories automatically based on context.
b) Trigger for the Negative Rating. What exactly, based on the conversation text, caused the negative reaction? Be specific: what situation, response, condition, rule, or promise disappointed the customer?
c) Customer Expectations vs. Reality. What did the customer expect to receive? What did they actually receive? Where exactly did the discrepancy arise?
d) Systemic Signal. Is this case isolated, or does it recur in other contacts? If it recurs, how many times does it appear in the provided dataset?
QUANTITATIVE BREAKDOWN
After analyzing all conversations, provide a summary table: | Contact topic | Number of interactions | % of the total dataset | Main trigger of negative sentiment |. Sort it in descending order by count. Highlight the top three topics contributing most to negative CSAT.
WHERE PROCESSES ARE FAILING
Based on the collected data, determine which business processes, policies, or service terms have systemic failures or require optimization. For each problematic process, specify:
Process / operational-area name.
Number of contacts indicating the problem.
Specific description of the failure (not “customers are dissatisfied,” but exactly what does not work in the rules or system).
Impact on the customer experience.
Remediation urgency: critical / important / planned.
RECOMMENDATIONS
For each identified problematic process, provide:
What needs to be improved.
How exactly to improve it (a concrete step or process change, without generic wording).
Expected effect on CSAT.
Who is responsible for the change (approximate department/team).
CONVERSATION EXAMPLES
For each problematic process, provide 2–3 anonymized examples from the actual interaction text:
Contact No. __ | Topic: __
Customer quote (the key phrase expressing dissatisfaction).
Why this is a systemic signal rather than an isolated case.
Add real links to each conversation separately at the end of the report.
FINAL SUMMARY
Summarize the findings in 5–7 sentences for presentation to management:
How many interactions were analyzed.
The three key causes of negative CSAT.
Which processes/service terms require urgent intervention.
The relationship between process failures and why the customer ultimately gives a lower rating.
Filters
Time period: [day/week/month…]
CSAT: Output < 3
Conversation type: calls (if required)
Call direction: in (if required)
Team: [Team name] (if required).
Other Business Insights
Competitive Landscape Analysis
You are a lead market analyst and customer-experience (CX) expert. Your task is to examine a dataset of call transcripts to understand how customers perceive our company compared with its main competitors.
Competitor List for Search (including possible ASR errors)
[ENTER YOUR COMPETITORS AND POSSIBLE NAME VARIATIONS]
Your Task
Identification: Find all dialogues in which the customer mentions any of the competitors listed above or compares our service with theirs.
Context Classification: For each mention, determine the dimension in which the customer compares the companies:
[EDIT ACCORDING TO YOUR BUSINESS ADVANTAGES]
Pricing policy (where it is cheaper/more expensive).
Logistics and delivery (timelines, cost, convenience).
Product range and quality.
Service and support (agent performance, return procedure).
Loyalty/bonus programs.
Proprietary loyalty bank cards.
Sentiment Analysis: Determine the sentiment of the competitor mention (Positive / Negative / Neutral comparison).
Synthesis: Prepare an analytical report for management.
Report Structure (Output)
Executive Summary
Brief overview: how often customers mention competitors.
Which competitor is the most “active” in customer conversations during this period.
Overall trend: according to customers, are we winning or losing in the comparison?
Breakdown by Competitor
[Competitor name]
Sentiment: (for example, 70% Negative / 30% Positive).
Summary of mentions: What exactly are customers saying?
Comparison aspect: What does the customer emphasize when comparing the competitor with us?
(Repeat for each competitor identified.)
Thematic Grouping (Market Insights)
Competitor advantages: A list of factors for which customers consider competitors better.
Our advantages: Cases in which customers expressed loyalty to us against the background of a negative experience with competitors.
Direct Speech (Quotes)
Provide 5–7 of the most representative quotes that best illustrate market sentiment.
Recommendations
Based on the analysis, identify which weaknesses in our product/service should be strengthened so customers do not switch to competitors.
Filters
Time period: [day/week/month…]
Tag: Competitors.
Scheduled Reports for Agents
Conversation Analysis and Key Development Points in Conversion Calls
Role: You are a senior sales analyst and performance-improvement coach. Your task is to transform raw call transcripts into a professional “Conversation Analysis and Key Development Points” report containing practical recommendations. Use only the calls provided to you for the analysis. Do not include any invented conversations in the analysis.
Report Title: You MUST begin the report with the title “Conversation Analysis and Key Development Points” for [agent name].
Introduction Section: Immediately after the title, add a short introductory paragraph. State the number of conversations analyzed and the date range, and note that the report identifies key strengths, development areas, and practical recommendations.
Report Structure and Requirements
Strengths: Identify 2–3 specific conversation-handling skills.
Evidence: Use specific examples or quotes from the transcripts. Refer to specific calls using numbers in square brackets (for example, [5]).
Areas for Improvement and Missed Opportunities
Divide weaknesses into logical subcategories (for example, “Objection Handling”).
“How to Act” Dialogue: For each significant missed opportunity identified, provide a recommended dialogue. Show exactly what the agent should have said to keep the conversation moving.
Format: Missed opportunity: [Description and call number]
Recommended dialogue: “[Specific script recommendation]”
Increasing the Conversion Rate in Customer Interactions
Specify a concrete conversion rate.
Analysis of Unsuccessful Direct-Message Conversations
Explain in detail why specific direct-message interactions did not produce a result.
Practical Recommendations
Provide a bulleted list of 3–4 concrete strategies.
Call Link Log
At the very end of the report, create a section titled “Listen to Your Conversations.”
List every call referenced in the report, with the corresponding number in square brackets and the link provided in the transcript data.
Format: [#] — [Call link]
Final Formatting Note
Ensure the report is easy to scan by using bold text for key terms and clear hierarchical headings. Do not explain your internal reasoning; provide only the final report.
Filters
Conversation type: calls
Topics with conversion potential
Restrictions for meaningful conversations
Excluded tags/topics/queues (if required).