Once you have successfully connected with Zendesk, you can create your first project by importing your Zendesk tickets. Click on Create Project > Start from Zendesk to get started.

Lang.ai will replicate the tags you select (based on your form fields) so you can start creating automations as soon as possible.

We will look for patterns in order to create transparent tags that you will be able to edit later on. For example, if you have a tag called "login_issues", the algorithm will analyze all the tickets with that tag in order to find patterns to learn how these kind of tickets look like. Once finished, it will create a tag with the same name in your Lang.ai project, but this time it will be able to classify the new tickets automatically.

Once you have imported your tickets it will take between 5 minutes and some hours (depending the amount of tickets) to be processed before it suggests you tags that you can use for automated tagging. You can see how you can start using tags in Lang.ai in this article.

By default, Lang.ai is configured to import the last three months of tickets that were created vie the email and web channels. You can change this query to include the tickets that represents your support interactions in the best way possible. These imported tickets will be used to replicated the tags you already have and to create new ones. In a few words, they will be used to create your classifier.

Setting a tickets query

You can copy a query that you already have from your Zendesk account in the Tickets > Search section. An example query may be:

Once you have your tickets query, click on Show Advanced Options and paste it in the Zendesk Tickets Query field.