Skip to content

Post-Resolution Knowledge Base Updater

Example prompt: "After a support ticket is resolved, summarise the issue and solution into a draft knowledge base article in Notion. If the same question has come up 3 or more times this month, flag it in #kb-updates on Slack so we can publish it."

How to automate knowledge base updates with GloriaMundo

The Problem

Most support teams know they should keep their knowledge base current, but the work of writing articles falls to the bottom of the priority list. Resolved tickets contain perfectly good answers to real customer questions, yet that knowledge stays locked in individual ticket threads. The same questions get answered from scratch, week after week, because nobody has time to turn the answers into public documentation.

How GloriaMundo Solves It

We build a workflow that runs whenever a ticket is marked as resolved. An LLM step reads the ticket conversation and distils it into a structured knowledge base article — a clear title, a description of the problem, and the step-by-step solution. A state step tracks how many times each question topic has appeared this month. If the count crosses a threshold you set (say, three occurrences), a Slack notification alerts your documentation team that this topic deserves a published article. The draft is saved to Notion, ready for a quick review and polish before publishing. Glass Box preview shows you the generated article and the recurrence count, so you can adjust the threshold and article format before enabling the workflow.

Example Workflow Steps

  1. Trigger (webhook): Fires when a support ticket status changes to "resolved".
  2. Step 1 (integration): Fetch the full ticket conversation — customer messages, agent replies, and resolution notes.
  3. Step 2 (LLM): Summarise the conversation into a knowledge base article with a title, problem description, and solution steps.
  4. Step 3 (state): Increment a counter for this question topic. Check whether it has exceeded the recurrence threshold.
  5. Step 4 (integration): Create a draft page in Notion with the generated article content, tagged with the topic category.
  6. Step 5 (conditional): If the recurrence threshold is met, notify the documentation team.
  7. Step 6 (integration): Post a message to #kb-updates on Slack with the topic, recurrence count, and a link to the Notion draft.

Integrations Used

  • Notion — stores draft knowledge base articles for review
  • Slack — notifies the documentation team about recurring topics ready for publishing

Who This Is For

Support leads and documentation managers at teams where the knowledge base is always out of date and agents keep answering the same questions manually.

Time & Cost Saved

Writing a knowledge base article from scratch typically takes 20-30 minutes per topic. If your team resolves 5 recurring topics per month that should become articles, that is 2-3 hours of documentation work. This workflow produces a solid first draft automatically, reducing the effort to a quick review and edit — roughly 5 minutes per article. Over a month, you save around 2 hours of writing time and, more importantly, customers start finding answers before they need to open a ticket. The workflow uses LLM, state, and integration steps, costing a few credits per resolved ticket.