Skip to content

Ticket Response Drafter

Example prompt: "When a new support email comes in, draft a helpful reply based on our FAQ docs and similar past tickets, and save it as a Gmail draft for me to review."

How to automate support ticket responses with GloriaMundo

The Problem

Every support ticket requires the same sequence: read the customer's message, understand the issue, look up the relevant documentation or past resolutions, and compose a thoughtful reply. Even experienced agents spend 5-10 minutes per ticket on this process, and for less common issues the research phase can take much longer. The quality of responses also varies depending on who handles the ticket — one agent might reference the exact help article, while another gives a vague answer from memory. The knowledge to answer most tickets already exists across your FAQ documents and historical resolutions; the bottleneck is retrieval and composition.

How GloriaMundo Solves It

We build a workflow that triggers when a new support email arrives. An integration step fetches the email content — the customer's question, any error details, and their account context. An LLM step reads the incoming message against your knowledge base documents and FAQ content, identifies the most relevant information, and drafts a personalised reply that addresses the specific question with accurate details. It references the appropriate help articles and uses a tone consistent with your team's style. Rather than sending the reply automatically, the workflow saves it as a Gmail draft so an agent can review, adjust, and send it with confidence. A notification is posted to Slack so the team knows a draft is ready for review. Glass Box preview shows you the drafted reply and the knowledge base sources it drew from before anything is saved, so you can check the quality and relevance.

Example Workflow Steps

  1. Trigger (webhook): Fires when a new support email arrives in the team inbox.
  2. Step 1 (integration): Fetch the email content from Gmail — sender, subject, body, and any attachments or error details.
  3. Step 2 (llm): Analyse the customer's issue against the knowledge base and FAQ documentation. Draft a personalised reply that addresses the specific problem, references relevant help articles, and matches the team's communication style.
  4. Step 3 (integration): Save the drafted reply as a Gmail draft in the support inbox, threaded under the original message.
  5. Step 4 (integration): Post a notification to #support on Slack with the ticket subject, a one-line summary of the issue, and a note that a draft reply is ready for review.

Integrations Used

  • Gmail — receives incoming support emails and stores the drafted replies for agent review
  • Slack — notifies the support team that a new ticket has a draft reply ready

Who This Is For

Support teams and customer-facing staff who handle a steady flow of inbound tickets and want to reduce the time between receiving a question and having a quality reply ready. Especially useful for teams where knowledge is spread across multiple documents and newer agents spend significant time searching for the right information.

Time & Cost Saved

Drafting a support reply manually — reading the ticket, finding relevant documentation, composing the response — takes 5-10 minutes per ticket. For a team handling 30-50 tickets per day, that is 2.5-8 hours of drafting work daily. This workflow delivers a ready-to-review draft within seconds of each ticket arriving, cutting per-ticket handling time by 50-70%. Over a month, that is 50-160 hours of agent time redirected from composition to review and more complex issues. The workflow uses integration and LLM steps, costing a few credits per ticket.