To Start Installing the Integration;
To install the MonkeyLearn Ticket Tagging integration for Zendesk, go to the page in the Zendesk Marketplace and click install. Once installed you will be taken through the following steps.
1) Grant MonkeyLearn permissions to read and write to Zendesk tickets. MonkeyLearn will only read the title and descriptions of tickets and will only write tags to the fields selected by the user (in step 3). No other information will be accessed or changed.
2) Insert your MonkeyLearn API Key. You must have an account with MonkeyLearn to use this integration. Your API key will connect the integration with your MonkeyLearn account and give you access to both your text analysis models and publicly available models. Any queries that are consumed via the integration will be counted against this account.
3) Choose the Ticket Fields to Tag Automatically by selecting a model from your MonkeyLearn account in the left column, and then selecting the ticket field you would like that model to output to in the right column. You may select a previously created ticket field, or you may create a new one.
4) Select the Automation Confidence Level you would like to apply to the tagging automation. With this setting, you can choose to tag only the tickets with the most confident results. The higher the confidence level, the more tags will be applied.
5) Activate the integration. With the integration activated, MonkeyLearn will automatically begin to tag incoming tickets according to the configurations made previously.
Changing Settings for the Integration
Click on the MonkeyLearn Logo on the left bar of the Zendesk App to open up the settings options.
Under Settings, you can make changes to the ticket fields and models from MonkeyLearn that have been configured to automatically tag incoming tickets. You can also change the confidence level you want to operate with.
Advanced FAQ and Modifications for the Integration
How does the integration work?
This integration creates a Zendesk trigger that sends the title and description of every new ticket to MonkeyLearn. The text is then analyzed by the models selected and a result is returned to Zendesk.
Can I select which tickets to automatically tag?
You can modify the MonkeyLearn trigger by going to your Zendesk settings. Look for Settings > Triggers and select MonkeyLearn New Ticket Trigger. There you can add new conditions to the trigger. For example, you may want to tag tickets for only a given group in Zendesk.
Does this integration use my MonkeyLearn queries?
Yes. Each time a model receives a text, it counts as a query. Please see your account in MonkeyLearn for your current usage and account limit.
What happens if I run out of queries?
If you have reached your query limit, new tickets will not be tagged. See our plans for upgrade options or talk to our team.
Are Chats in Zendesk also considered?
No, chats are not accessed as part of this integration.
How are new Ticket Fields created by the integration?
When a MonkeyLearn model (either classifier or extractor) is selected, you have the option to create a new field to receive the tags from that model. If a classifier is selected, the field that is created will be a multi-select field with options that exactly match the tags in the model. If an extractor is selected, a new text field will be created.
Can I change my MonkeyLearn token?
It is not possible to change your MonkeyLearn token. To use a new token, you can delete the current integration and reinstall it using a new token.
What are classifiers and extractors?
MonkeyLearn offers two kinds of text analysis models:
Classifiers classify texts with a tag (or category) that is usually not present within the text, so a classification has to be deduced by the model based on the text provided. Sentiment Analysis is an example.
Extractors extract text or data points from a given ticket, as the result is found within the text. Keyword Analysis is an example.
What kind of fields and outputs can I use with MonkeyLearn?
Extractors can be used to automate Text and Multi-line fields. All extractions will be joined by commas.
Classifiers can be used to populate Drop-down, Multi-select, Text, and Multi-line fields. Checkbox, Numeric, Decimal, Date, Credit card, and Regex fields are not supported.
In the case of Drop-down and Multi-select, their tag values in Zendesk must match the tags in MonkeyLearn, otherwise, the result will be ignored. In Multi-select fields, every MonkeyLearn tag that matches the selected confidence level will be applied to the field. Meanwhile, in Drop-downs only the most confident value will be used. When classifiers are used with Text or multi-line fields, all the tags that match the selected confidence level will be joined by commas.
What can I do if I can't match a Drop-down field with a Zendesk field?
MonkeyLearn integration actually analyzes the tags, not the names, of the options of the Zendesk fields. Those tags are generated automatically by Zendesk when the options of the Drop-down (or Multi-select) fields are created, but they are not automatically updated when the name is changed. Look for Settings > Ticket Fields and select the field you are trying to match. Then, check the option Show tags: those are the values that should match with the MonkeyLearn classifier tags.
Questions or Support for the MonkeyLearn Ticket Tagging Integration?
Please contact us at firstname.lastname@example.org.