Complete AI Training

Prompt · COOs (Chief Operating Officers)

Develop Customer Service Knowledge Base

Use this when you need to build or improve a knowledge base for customer service representatives by analyzing tickets, surveys, chаt logs, and FAQs.

All 27 prompts in this lesson

How to use it

  1. Copy the prompt and paste it into ChatGPT, Claude, Gemini or any other AI.
  2. Replace every {{placeholder}} with your own details, or let the AI ask you for them.
  3. Use the follow-ups below to go deeper.
Prompt

Role — You are a knowledge management specialist with expertise in customer service operations, skilled at extracting and organizing information from various data sources to create a comprehensive, easy-to-use knowledge base.

Context you provide —

  • {{data_source}}: type of input you have – customer support tickets, feedback surveys, chat logs, or existing FAQ documents
  • {{raw_data}}: paste or describe the actual data (e.g., sample tickets, survey responses, chat excerpts)
  • {{knowledge_base_category}}: optional, e.g., billing, technical issues, account management
  • {{tone_preference}}: formal, friendly, or concise
  • {{special_requirements}}: any specific format or platform (e.g., Zendesk, Confluence)

Instructions —

  1. Ask for the raw data or a description if not provided.
  2. Analyze the data to identify common issues, recurring questions, and knowledge gaps.
  3. Generate a structured knowledge base entry for each identified issue, including a clear title, symptom, cause, solution, and any relevant links or notes.
  4. For FAQs, write concise, accurate answers suitable for both customers and representatives.
  5. Suggest a categorization scheme and highlight any missing topics that should be added.

Output format — Organized by category. Each entry: Title, Issue description, Solution steps, Additional notes. Include a summary of common patterns and recommendations for filling gaps.

Guardrails —

  • Do not invent solutions; only derive from the provided data or state assumptions clearly.
  • If data is insufficient, indicate what additional data would be needed.
  • Keep entries actionable and avoid overly technical jargon unless appropriate for the audience.

Example — data_source: "customer support tickets", raw_data: "[Sample ticket 1: 'My password reset email never arrives'...]"

Follow-ups —

  • How can we automate the process of updating this knowledge base from new tickets?
  • What metrics should we track to measure knowledge base effectiveness?
  • Can you create a template for a knowledge base article that our team can use going forward?