Implementing ticket automation for advanced AI agents

In AI agents - Advanced, the AI-powered Ticket Automation feature eliminates repetitive and mundane support ticket tasks so your agents can focus on complex tickets and providing the best customer experience.

Ticket Automation is able to:

• Add tags and internal notes to Zendesk Support tickets automatically
• Create personalized messages from user info 
• Update ticket information 
• Merge tickets from the same requester
• Routing tickets to agents

Step 1: Import historical data

• This is quick and easy if you already use Zendesk Support. See importing data from Zendesk Support. If you don't, you can also import data manually. 
• If you don't have historical data, see What to do if I don't have any historical data?

Step 2: Complete Preparation

Your AI agent in Zendesk Support needs its own account. For details see prepare connecting your email AI agent to Zendesk Support.

Step 3: Writing time!

Think about the replies you would like to use for each intent. You can use markdown to style the text replies.

• It's good to keep those inline with your brand, tone of voice and customer persona
• Have a look at the actions available for ticket automation to support your customer support processes

Step 4: Connect and test your email AI agent

Make sure the preparation for connecting your AI agent to Zendesk Support is done before connecting your AI agent. 

Points to keep in mind when testing:

• Don't troll the AI agent
• Ask the same question in different ways  
• Ask test participants to take screenshots to accompany their feedback
• Gather feedback in one place - you can easily find all the records in Conversation Logs but gathering feedback and screenshots in one place will makes iterating more efficient for you
• If you have set up actions and internal notes, see if they are inline with your current practice. 
  • If the customer support processes will be changed with the launch of the AI agent, this is a good time to take screenshots so your agents know what to expect.

Step 5: Iterate

With all the feedback you've gathered, time to make changes to your AI agent. It could be a different copy for the reply or adding actions to add tags and merge tickets to reduce your backlog.  

Step 6: Launch and maintain

After launching the AI agent, meaning having an AI agent working in your team, keep an eye on it and see how it's performing. Keep training the AI agent and make adjustments to your copy as you go in order to make sure your AI agent always sound and act smart!