This chapter provides information on how to configure AI Ticket Assistant.
Thanks to the AI Ticket Assistant (ATA) configuration you can short-cut issue reporting by analysing the user’s natural-language description, proposing the correct request types, and autofilling the form’s fields.
In the table below you can find details concerning the available Assistant’s functionalities.
| Functionality | Details |
|---|---|
| Greeting | Widget follows the customer’s profile language and greets them accordingly. |
| Intent detection | ATA analyses the problem described by the user in a plain language and suggests the right request type based on the findings. |
| Data capture | ATA parses the free-text description, paraphrases it into Summary or Description and extracts field values. |
| Proactive follow-ups | If some data is missing, the bot politely asks focused questions, encouraging the user to provide more details for broader context. |
| Field validation | When a required field is unsupported, ATA shows a fallback message and asks the reporter to fill it manually later. |
| Preview and submit | Users can review and override any suggested value before the ticket is created in JSM. |
Before you start using AI Ticket Assistant, learn about the app installation and configuration.
Once the app is successfully installed, you can configure it in Manage Apps > AI Ticket Assistant > Configuration.
If you’re using the app for the first time after installation, the Assistant may need a few moments to fully sync your Jira configuration. This ensures correct setup to provide you with the best experience.
The configuration screen can be accessed by navigating to Administration > AI Ticket Assistant > Configuration. The view is the same for both app editions (Standard and Advanced) unless noted otherwise.
| UI Element | Functionality | Usage scope |
|---|---|---|
| Global toggle | Show AI Ticket Assistant on Customer Portal turns the entire app on or off for every portal | Acts as a quick kill-switch if you need to disable ATA company-wide (for example, during maintenance or after reaching the credit limit). |
| Tabs | Access control and Usage & data | Splits day-to-day administration (like determining ATA visibility) from operational statics (for example, obtaining data related to conversations or tokens). |
To start using the Assistant, navigate to Manage Apps > AI Ticket Assistant > Configuration and switch the Show AI Ticket Assistant on Customer Portal toggle to turn the Assistant on so that it’s visible to the Customer Portal users. To turn it off, use the same toggle.
In the Standard app edition, by accessing Configuration, you also need to provide your OpenAI API key for the app validation. To learn more about each edition of the app, navigate to the App editions chapter.
The Access control view is the same for both app editions (Standard and Advanced). It allows you to decide about who can use the Assistant and determine portals for which the Assistant should be available.
First, choose where the Assistant appears:
Next, choose who can use the Assistant:
Click Save changes to apply your settings or Discard to cancel and revert any unsaved changes.
The Usage & data tab gives administrators an up-to-the-minute view of how much the Assistant is being used in the current billing cycle. The tab view is different for the Standard and Advanced app editions. The AI Ticket Assistant provides users with the ability to manage and optimize their token usage costs to best fit their workflow and budget as each edition offers a different level of flexibility and cost control.
The table below presents details on differences oh the Usage and data view between available depending on the app edition.
| Edition | View details | Information provided |
|---|---|---|
| Standard | - Conversations with the Assistant presenting chat sessions started so far. -Usage presenting total tokens consumed. |
Shows exactly how many chats happened and how many OpenAI tokens they cost. |
| Advanced | - Conversations with the Assistant presenting chat sessions started so far. -Credit usage presenting the progress bar and exact percentage of tokens usage (Green < 70 %, Yellow 70 - 90 %, Red > 90%). |
Let’s you see at a glance whether usage is within limits or about to exceed it. |
To learn more about each edition of the app, navigate to the App editions chapter.
In the Standard edition you need to provide your own OpenAI API key and get billed directly by OpenAI so the Usage & data tab won’t provide you with the percentage of your monthly data limits.
What you can see is the number of sessions for the current billing cycle and the amount of consumed tokens.
The Advanced app edition includes OpenAI access managed by the app and presents usage covered in your subscription. That’s why, you can easily track and monitor your current usage with the Data & usage view where detailed statistics and limits are displayed.
This section displays the number of sessions (interactions between users and the AI assistant) for the current billing cycle.
The Usage section shows the percentage (%) of your monthly token limit that has been used. It resets automatically at the beginning of each billing cycle.
Once the limit is exceeded, ATA automatically disables itself globally. The application stays inactive until the quota resets at the start of the next billing period or when additional credits are purchased with the help of the Deviniti team.
To use the Assistant efficiently, consider the following:
The Assistant automatically synchronizes your Jira data every 24 hours to ensure it stays up to date.
If recent changes aren’t visible, you can manually update the Assistant’ index. Also, do this after adding changes to your Portals, request types, permissions or Confluence Knowledge base.
The Last updated section shows the timestamp of the last successful sync.
This section explains what happens when AI Ticket Assistant (ATA) stops working due to one of the following reasons:
To check details on different editions of AI Ticket Assistant (ATA), navigate to the App editions chapter.
When the monthly token quota is fully consumed, the application automatically deactivates to prevent further usage.
The table below presents you with details on the app deactivation.
| Area | Result | Details |
|---|---|---|
| Service shutdown | ATA disables itself globally. | All portals are affected immediately. |
| Customer Portal | The chat widget disappears. | End-users (without admin or agent permissions) can’t start or continue conversations. The widget is completely hidden. |
| Customer Portal | The warning banner appears. | The banner informs you that the token limit has been exceeded. |
The application remains inactive until:
To prevent deactivation, best practices for administrators include:
Regularly check the Usage & Data tab. The colour of the usage bar gives an instant status update.
If usage approaches 80%, consider disabling ATA on less-critical portals via Access Control.
Encourage end-users to write concise messages because shorter inputs consume fewer tokens.
When the OpenAI API key provided by the administrator becomes invalid or inactive, the application remains enabled but chat conversations can’t start.
The table below presents you with details on the invalid API Key scenario.
| Area | Result | Details |
|---|---|---|
| Customer portal | The chat widget remains visible and active. | End-users can click Start conversation, but as soon as they send the first message, they see an error message. |
| Administration panel | A red error notification appears next to the API key field. | All toggles and settings remain enabled and ATA isn’t automatically disabled. |
| Service behavior | Conversations fail to start until a valid key is provided. | Once a correct key is saved, the chat resumes immediately and no additional steps are required. |
You can update the API key at any time in the Configuration tab. The system instantly validates the new key and restores functionality if the key is correct.
Back to App page