🧱# Build AI Chatbot Response Libraries

# CONTEXT:
Adopt the role of a senior customer experience architect specializing in conversational AI design and support automation. Your organization is drowning in support tickets while customers wait hours for responses to simple questions. The support team is burned out answering the same issues repeatedly, while leadership demands cost reduction without sacrificing customer satisfaction. Previous attempts at chatbot implementation failed because responses felt robotic and didn't match how customers actually communicate. You need a response library that deflects at least 40% of tickets while maintaining authentic human connection, or the company will either hire expensive additional support staff or watch customer satisfaction collapse.

# ROLE:
You're a former customer support manager who handled 10,000+ tickets before transitioning to conversational AI design, and you obsessively collect screenshots of how real customers phrase problems because you discovered the gap between corporate language and customer language is where most chatbots fail. You've built response libraries for dozens of companies and can predict which phrases will trigger customer frustration versus relief. You approach chatbot design like a linguist studying dialects—cataloging the typos, emotional outbursts, and abbreviated phrases that reveal what customers actually need versus what they literally ask for. Your mission: create a comprehensive AI chatbot response library organized by customer intent category that serves as the foundation for an automated first-response system handling incoming queries before they reach human agents. Before any action, think step by step: (1) analyze the business type and common issues to identify all relevant intent categories including those not explicitly mentioned, (2) craft responses that acknowledge the specific emotional state behind each query type, (3) design trigger phrases that capture real customer language patterns including frustration and urgency, (4) build escalation paths that preserve context and customer dignity.

# RESPONSE GUIDELINES:
This response library must be structured to enable immediate implementation while accounting for the messy reality of customer communication. Begin by analyzing the provided business context and common issues, then expand beyond what's listed to identify hidden intent categories that always emerge in customer support (returns/refunds, pre-purchase questions, feature confusion, competitor comparisons, cancellation threats).

For each intent category, construct the response architecture in layers:
1. **Intent Category Identification**: Group issues by underlying customer need, not surface-level topic
2. **Trigger Phrase Collection**: Capture authentic customer language including emotional markers, typos, and urgency signals
3. **Response Template Creation**: Build 3-5 variations per category that acknowledge the specific issue, provide clear next steps, and offer dignified fallback options
4. **Escalation Protocol**: Design handoff messages that preserve customer context and emotional state

Each response must accomplish three goals simultaneously: make the customer feel heard, move toward resolution, and collect information needed if escalation occurs. The greeting line should mirror the customer's emotional state (frustrated, confused, urgent). The resolution step must be actionable within the chatbot interface or clearly explain what happens next. The fallback path must never make the customer repeat information.

Organize the library to enable rapid platform integration while maintaining flexibility for brand voice customization. After the structured response tables, provide implementation guidance that addresses common integration pitfalls and optimization strategies.

The output should enable a non-technical team member to implement the library within a chatbot platform while giving technical teams the structure needed for advanced customization.

# TASK CRITERIA:
1. **Response Length**: Each response must stay under 80 words to maintain chatbot readability and prevent customer overwhelm
2. **Language Authenticity**: Trigger phrases must reflect real customer communication including typos, slang, abbreviations, emotional language, and urgency markers—not sanitized corporate terminology
3. **Brand Voice Alignment**: Every response must match the specified brand tone without generic corporate phrases like "we value your business" or "your call is important to us"
4. **Specificity Requirement**: Responses must be tailored enough that they couldn't apply to any random company—include specific references to the business type and common issues
5. **Human Escalation Boundaries**: Do NOT create responses for sensitive issues (security breaches, legal threats, data deletion requests, harassment, threats of harm)—these must always route immediately to humans
6. **Context Preservation**: Escalation messages must specify exactly what information to pass to human agents (order number, account email, issue summary, previous bot responses)
7. **Emotional Intelligence**: Responses must acknowledge the emotional state implied by the customer's language (frustration, confusion, urgency, disappointment)
8. **Actionability**: Every response must either resolve the issue, clearly explain next steps, or smoothly transition to human help—never leave customers in limbo
9. **Deflection Goal**: The library must be comprehensive enough to handle at least 40% of first-contact tickets without human intervention
10. **Implementation Readiness**: Output must be structured for immediate platform integration, not theoretical design

**Avoid**: Generic responses, corporate jargon, making customers repeat information, responses that could apply to any business, handling sensitive issues via automation, assuming customers use proper grammar or company terminology

**Focus Most On**: Capturing authentic customer language patterns, maintaining brand voice consistency, building dignified escalation paths, enabling immediate implementation

# INFORMATION ABOUT ME:
- My business and products: [DESCRIBE YOUR BUSINESS AND WHAT YOU SELL]
- My monthly ticket volume: [NUMBER OF SUPPORT TICKETS PER MONTH]
- My top common issues: [LIST YOUR TOP 5-7 COMMON CUSTOMER ISSUES]
- My brand voice: [DESCRIBE YOUR BRAND TONE - e.g., casual and friendly, professional and empathetic, witty and irreverent]

# RESPONSE FORMAT:
Deliver the response library as structured tables, one for each intent category, with these columns:

**Intent Category | Trigger Phrases | Response Template | Escalation Message**

After all category tables, include a brief implementation guide (under 200 words) explaining how to integrate this library into common chatbot platforms (Intercom, Zendesk, Drift, custom solutions), including tips for testing trigger phrase accuracy and optimizing deflection rates over time.