AI Voice Agent Prompt Design
Budget: ₹750 – ₹1,250 INR
I’m putting together a voice-based AI assistant that will field technical support calls for consumer hardware devices. To move quickly, I need a specialist who can craft clear, structured prompts and dialog flows that let the agent:
• Understand and triage basic troubleshooting requests (no deep diagnostics—think power-cycle steps, cabling checks, simple configuration guidance).
• Ask follow-up questions that feel natural yet gather the key device details the first time.
• Deliver concise, jargon-free solutions or next-step instructions, plus recognize when to escalate to a live rep.
You’ll start with a short discovery session to review the product line and any existing knowledge base articles. From there, I’m expecting:
1. A conversational script outline covering greeting, authentication, issue identification, solution paths, and graceful hand-off.
2. Prompt templates (system, user, and assistant roles) optimized for a modern LLM-powered voice stack.
3. A quick reference table mapping the most common hardware issues to the exact prompt sequence you’ve designed.
4. Brief testing guidance so I can validate responses in my chosen voice platform.
Keep the scope lean and focused on this single technical-support use case—no multilingual variants or advanced diagnostics for now. If the initial rollout succeeds, we can discuss expanding to deeper troubleshooting or additional product categories.
• Understand and triage basic troubleshooting requests (no deep diagnostics—think power-cycle steps, cabling checks, simple configuration guidance).
• Ask follow-up questions that feel natural yet gather the key device details the first time.
• Deliver concise, jargon-free solutions or next-step instructions, plus recognize when to escalate to a live rep.
You’ll start with a short discovery session to review the product line and any existing knowledge base articles. From there, I’m expecting:
1. A conversational script outline covering greeting, authentication, issue identification, solution paths, and graceful hand-off.
2. Prompt templates (system, user, and assistant roles) optimized for a modern LLM-powered voice stack.
3. A quick reference table mapping the most common hardware issues to the exact prompt sequence you’ve designed.
4. Brief testing guidance so I can validate responses in my chosen voice platform.
Keep the scope lean and focused on this single technical-support use case—no multilingual variants or advanced diagnostics for now. If the initial rollout succeeds, we can discuss expanding to deeper troubleshooting or additional product categories.