Automate Customer Support with AI

This rantir template is designed to assist customer support teams by automating the resolution of long-standing and forgotten JIRA issues, thereby improving team efficiency and issue management.

How it works

 
• A scheduled trigger runs daily to check for unresolved issues that have been open for a long time and imports them into the workflow.
 
• Each issue is processed in parallel as a separate subworkflow using the execute workflow node.
 
• A report is generated from the issue’s comment history, and AI is used to classify the issue, determining its current state and progress.
 
• If the issue is resolved, sentiment analysis is performed to assess customer satisfaction. If the sentiment is negative, a Slack message is sent to escalate the issue; otherwise, the issue is automatically closed.
 
• If no response has been initiated, an AI agent will search for and attempt to resolve the issue using similar resolved issues or information from the Notion database. If a solution is found, it is posted to the issue, and the issue is closed.
 
• If the issue is blocked and waiting for responses, a reminder message is added to prompt action.

How to use

 
• This template searches for JIRA issues that are older than 7 days and are not in the "Done" status. Ensure there are issues that meet this criteria, or adjust the search query accordingly.
 
• It works best if you frequently have long-lived issues that need resolving.
 
• Ensure the Notion tool is configured to avoid reading documents you didn’t intend to, such as private or internal documentation.

Requirements

 
• JIRA for issue management
 
• OpenAI for LLM
 
• Slack for notifications

Customizing this workflow

 
• You can adapt this workflow to classify issues as they are created. For example, it could be used for quality control tasks, such as ensuring reporting criteria are adhered to, summarizing and rephrasing issues for easier readability, or adjusting priorities.