Freshworks Prompts

Act as a Freshdesk supervisor and draft a standard operating procedure that teaches agents on team [TEAM] how to use Freddy AI Copilot's ticket summary generator on ticket [TICKET_ID] before their first reply. Include when to read the full thread instead, how to edit the AI draft, and how to keep the customer's own words in the final reply so summaries never replace human judgment.

Review the Freshdesk solution article titled [ARTICLE_TITLE] and rewrite it for Freddy AI Agent retrieval. Add a direct answer in the first two lines, short headings, step by step fixes, and 3 to 5 natural customer phrasings of the question. Keep the tone helpful and neutral, remove outdated references, and end with a clear escalation path to a human agent for unresolved cases.

Write a Freshdesk Self Service help article that deflects repeat tickets about [TOPIC]. Start with the most common question, give the shortest working answer first, then expandable troubleshooting steps. Add 5 real customer search phrases, link two related articles, and add a contact trigger when the fix fails. Keep sentences short so the Freddy AI Agent can quote it verbatim.

Design Freshdesk auto triage and routing rules for [DEPARTMENT]. Define how Freddy AI should set priority, group, and type for tickets matching keywords [KEYWORD_1], [KEYWORD_2], [KEYWORD_3]. Include one escalation rule for VIP customers tagged [TAG], one SLA policy for urgent tickets, and a fallback rule that routes unclassified tickets to [FALLBACK_QUEUE] for manual review.

Build a ticket QA scorecard for Freshdesk agents handling [PRODUCT_LINE] tickets. Score tone (empathy, clarity), accuracy (correct solution article used), process (SLA met, fields filled), and Freddy Copilot usage (summary checked, reply suggester edited before sending). Each criterion gets 1 to 5 points with one example of good and poor behavior, plus a coaching note template for low scores.

Write a Freshdesk CSAT recovery workflow for ticket [TICKET_ID] rated [SCORE] with customer comment "[COMMENT]". Include a same-day apology template, one question to clarify the root cause, an owner assignment rule, a Freddy Insights tag to track the failure reason, and a 7-day follow-up message. Keep every message under 80 words and require a human to approve before sending.

Create 3 Freshdesk canned responses for [SCENARIO]: one first reply, one status update, and one resolution confirmation. Each template uses placeholders [CUSTOMER_NAME], [TICKET_ID], and [AGENT_NAME], stays under 120 words, and includes a placeholder for the relevant solution article link. Add a usage note explaining when Freddy Copilot's canned response suggester should trigger each one.

Design a Freddy Insights weekly report for support lead [MANAGER_NAME] covering ticket volume, first response time, resolution time, CSAT, and Freddy AI Agent deflection rate. For each metric add target versus actual, week over week change, one likely cause from ticket tags, and one recommended action. End with the top 3 emerging topics that need new solution articles.

Write 5 training intents for the Freddy AI Agent Studio bot answering questions about [PRODUCT]. For each intent give the intent name, 6 sample user utterances in natural language, the correct answer drawn from solution article [ARTICLE_URL], one follow-up question the bot should ask, and the fallback action when confidence is low. Keep answers under 60 words each.

Write a 30-day onboarding playbook for new Freshdesk agents on team [TEAM]. Week 1 covers ticket views, canned responses, and SLA rules. Week 2 covers Freddy AI Copilot features like reply suggester and summary generator. Week 3 covers shadowing with a buddy and first solo tickets. Week 4 covers QA review and targets for first response time, CSAT, and resolution rate.