What This Workflow Does
This workflow checks for unresolved JIRA tickets open more than 7 days.
It reads all conversations on each ticket and uses AI to decide the ticket’s status.
It tries to solve tickets automatically or asks for more details.
It sends reminders or closes tickets based on results.
This helps support teams save time and keep customers happier.
Tools and Services Used
- n8n: Automation platform to run the workflow.
- JIRA Software Cloud: Source of support tickets.
- OpenAI: Provides AI for state classification, sentiment analysis, and generating comments.
- Slack: Sends alerts for negative or unresolved tickets.
- Notion: Knowledge base for finding solutions.
Step-By-Step: How to Use This Workflow in n8n
Step 1: Import and Setup
- Download the workflow file by clicking the Download button on this page.
- Open n8n editor where you want to run this.
- Import the workflow using the Import from File option.
Step 2: Configure Credentials and IDs
- Add your JIRA API Key and connection in the workflow credentials.
- Set up OpenAI API Key for AI nodes (classification, sentiment, reply generation).
- Input Slack credentials and target Slack channel ID to get alerts.
- Connect Notion with API Key and update database or page IDs for knowledge base query.
- Check and update any IDs, email addresses, or URLs inside message nodes as needed.
Step 3: Test the Workflow
- Run the workflow manually on a test issue to see if it finds issues and processes AI steps.
- Check returned comments, classifications, and Slack messages.
Step 4: Activate for Production
- Enable the Schedule Trigger node to run daily automatically.
- Keep an eye on execution logs for errors or missing credentials.
- Adjust prompt texts or JQL queries if needed for tuning.
If running on your own server, consider using self-host n8n for better control.
Workflow Inputs, Processing, and Outputs
Inputs
- JIRA issues that are open and older than 7 days.
- All comments and metadata from those issues.
- Company knowledge base content from Notion.
Processing Steps
- Use JQL query in JIRA node to fetch unresolved tickets older than 7 days.
- Extract detailed metadata and gather all comments per ticket.
- Transform comments into easy readable text thread for AI.
- Use OpenAI GPT-4 model to classify ticket status (resolved, pending info, or no response).
- If resolved, run sentiment analysis on customer tone to decide next steps.
- Search similar tickets and product docs in Notion to try auto-solving unresolved tickets.
- Post AI-generated replies or reminders to tickets.
- Close tickets if solved or escalate negative sentiment tickets by sending Slack alerts.
Outputs
- Comment updates on JIRA tickets with AI responses.
- Automatic ticket closures or follow-up reminders.
- Slack notifications for unresolved issues with negative feedback.
Common Issues and Fixes
No Tickets Found by JIRA Node?
Check if the JQL in the Get List of Unresolved Long Lived Issues node is correct.
Try the same query manually inside JIRA search to see results.
AI Returns No Solution?
Make sure Notion knowledge base has enough info and is correctly connected.
Verify OpenAI API key is active and not over quota.
No Slack Messages?
Confirm Slack credentials and channel IDs are valid in n8n.
Test simple Slack messages before running automated alerts.
Customizations
- Adjust the “older than 7 days” filter in the JQL query to shorter or longer periods.
- Switch OpenAI models in AI nodes for balance between cost and quality.
- Change URLs in feedback request messages to your preferred review platform.
- Add more Slack notifications for events like reminders or escalations.
- Rewrite reminder prompts in the AI nodes to fit your company tone.
Summary of Benefits and Results
✓ Saves support team hours by automating checking and replying.
✓ Stops tickets from being forgotten or delayed.
✓ Improves customer happiness through faster info and feedback requests.
✓ Helps focus human attention on bad or complicated tickets by sending alerts.