AI Prompts for Customer Support -- 2

I’m building an automated customer-support flow and need a prompt-engineering specialist who can craft rock-solid instructions for a large-language model. The end product should let the bot handle our most common requests—answering FAQs, providing up-to-the-minute order-status updates, and guiding users through basic technical troubleshooting—without sounding robotic or slipping into inaccurate replies.

You’ll analyse the current support scripts, spot any knowledge gaps, then design concise, modular prompts that keep context tight and responses on-brand. A light copywriting touch is welcome when polishing the wording, but the heart of the job is the prompt logic itself: role definitions, user intents, guardrails, and fallback strategies that reduce hallucinations and escalation rates.

Deliverables
• A well-commented prompt library covering FAQs, order status, and tech troubleshooting
• A short implementation guide explaining variables, temperature settings, and token limits so my dev team can drop the prompts straight into our workflow
• Test transcripts demonstrating successful handling of at least ten real support scenarios for each category, with notes on any edge cases

Everything will be considered complete once the guide integrates cleanly with our existing LLM endpoint and achieves a minimum 90 % first-contact resolution in sandbox testing.