AI Categories use artificial intelligence to automatically classify customer support conversations into predefined topics. This powerful feature helps you route conversations to the right teams, track conversation types, and gain insights into customer needs without manual tagging.
Key benefits
Automated classification - Fin automatically categorizes conversations as they arrive.
Improved routing - Send conversations to your team inboxes based on AI-detected categories.
Better reporting - Track conversation volumes and trends across different topics.
Reduced manual work - Eliminate the need for teammates to manually tag every conversation.
Real-time insights - Get immediate visibility into conversation patterns and customer needs.
Note:
AI category detection is a beta feature available to selected customers only. If you're interested in joining the beta, please email Zoe at zoe.sinnott@intercom.io
Setting up AI category detection requires "Can manage automation settings" permission in your workspace.
How to set up AI Categories
Overview
Map out your topics.
Create a “AI category” Conversation Data Attributes (CvDA).
Create a second CvDA called “AI subcategory”.
For each CvDA, add descriptions to train Fin → think prompt engineering.
Add conditional attributes as needed.
Add the "AI category detection" action in your workflows.
Route conversations to the right team based on the AI Category.
Build a report to track AI Categories.
Step 1: Map out your topics
Before creating AI Categories, identify what types of conversations you need to track and route. It helps to plan your category structure externally in a spreadsheet first, then upload your options via CSV for efficiency.
Common category examples:
Topics - Billing, Account Management, Support, Technical
Sentiment - Positive, Negative, Neutral
Priority levels - Urgent, High, Normal, Low
Spam detection - Spam, Legitimate
Download the templates at the bottom of this article for AI Categories you can customize and upload as CSV files👇
Step 2: Create your AI Category attribute
Click Create attribute.
Click on the format dropdown and select List.
Fill in the "Name and "Description" fields for the attribute, e.g.
Name: "AI Category" Description: "Examply conversation categories".For data hygiene and better reporting, it's a good idea to make this a Required attribute that teammates must ensure is completed before closing a conversation.
Fill in the "List options" section with clear names and descriptions. You can click Add to add more category options.
Or, upload a CSV file where each option is on a separate row. The first column should contain the category option name, and the second column should include a description. Please note that the first row will be treated as column headers and will not be imported.
Note:
To use conversation data attributes for AI category detection, they must be in a list format and each option in the list must have both a name and a description.
The maximum number of list attributes per conversation data attribute is 250.
The maximum description length is 2,500 characters.
Step 3: Create AI Subcategory attribute optional
For more granular categorization, you can create conditional attributes where subcategories only appear when specific parent categories are selected.
Create a second conversation data attribute called "AI Subcategory".
Use the List format.
Upload a CSV file where each option is on a separate row. The first column should contain the subcategory option name, and the second column should include a description. Please note that the first row will be treated as column headers and will not be imported.
Or, add your subcategory options manually then save.
Now go back and edit your main AI Category attribute.
Click Conditions at the top.
Select an AI Category option, and then select the AI Subcategory attribute from the dropdown.
Choose the AI Subcategory options you want to be included.
Click Save when you've finished mapping subcategories to parent categories.
This is how it should look once you’re all set up:
Routing conversations with AI Categories
After creating a conversation data attribute with your AI Category options, you can start using it to route conversations automatically in workflows.
Enable AI category detection in a workflow
Go to Fin AI Agent > Workflows.
Create a new workflow or edit an existing one.
Add the AI category detection action anywhere in your workflow where you’d like categorize the conversation.
Select the AI Category attribute you created.
Now you can use the AI Category options within your workflow. For example, you can add a branching condition based on the AI Category detected and route the conversation to a specific team or teammate.
You can also add the AI detected category in a Note or a Message.
Save your workflow and then test the AI category detection using the workflow Preview.
Set your workflow live.
Tip: Workflows support conditional attributes, meaning AI category detection will respect any logic you’ve defined for when dependent attributes should be evaluated — and only apply values when relevant. Learn more about conditional attributes.
Set the AI category manually in a workflow
The AI category can be set manually in a workflow if necessary. This is particularly useful when using reply buttons to guide the customer selection and subsequent categories can be applied.
For example, in the workflow below, if the customer selects "Banners" as an option for the query then "Banners" is the category applied to the conversation.
It's also possible to utilize a combination of manual and automatic AI category detection in the same workflow if required.
Note:
It's not possible to manually remove an applied AI category in a workflow, you can only manually apply a new one to override the old one.
When using the “Set AI category detection” instead of the automatic AI category detection a note will not be left in the conversation.
View and override AI Category when needed
Teammates working in the inbox can see when a category has been detected by Fin in three places:
Conversation events within the conversation thread.
Internal notes.
The inbox sidebar (when conversation attributes are pinned).
Teammates can manually update AI Category values:
From the inbox - Update in the right sidebar.
Using macros - Create a macro and add an action to quickly change categories.
Before closing - Required attributes need to be completed before a conversation is closed.
Reporting and analytics
Filter reports by AI Categories
Use AI Categories to segment your reporting:
Add filters for "AI Category" or "AI Subcategory"
Create charts segmented by conversation type.
Track volume trends across different topics.
Monitor categorization accuracy
Review miscategorized conversations regularly.
Update category descriptions based on patterns.
Train teammates to flag incorrect categorizations.
Pro tip: Implement a feedback process where teammates can report categorization issues to continuously improve accuracy.
AI Category best practices
Adding clear names and detailed descriptions for each of your category options is the most critical step for accurate categorization.
Top tips:
Create clear, concise names - Choose short, descriptive names that immediately convey the category's purpose.
Write comprehensive descriptions - Take the time to write detailed descriptions and include all relevant information such as:
What conversations belong in this category.
Common issues or topics covered.
Relevant keywords customers might use.
Example customer questions.
Make categories distinct - Avoid creating categories that overlap too much. Your categories should be clearly different from each other, making it easy to determine which one best fits a given situation.
If Fin sets the wrong one, teammates can update the attribute value from the inbox.
Implement a process for teammates to flag miscategorization (iterate descriptions as gaps/issues are identified).
Example description for "Account Issues":
This category includes problems related to accessing, managing, or recovering a user's account. Common Issues: Login failures due to incorrect credentials, forgotten passwords, or system errors. Account recovery requests (e.g., "I forgot my password, how can I reset it?"). Username or email update requests. Issues with two-factor authentication (e.g., "I lost access to my authenticator app, how do I log in?"). Locked or suspended accounts due to security concerns. Problems with email verification or account activation. Keywords: login, password reset, locked account, 2FA, verification, recovery, authentication, username change.
Pro tip: Try passing your category names and descriptions to a writing tool such as Claude AI or ChatGPT to define them clearly.
Example: Write comprehensive descriptions for all of the topics listed - Include all relevant details about what belongs in the category. Think about every type of conversation that should fall under this category and describe them in the description. Providing a detailed description will help our AI Agent categorize support conversations correctly. Include keywords and examples of customer questions.
Real-world examples
Sentiment Categories:
Positive - A positive sentiment means the user who wrote the message seems generally happy or satisfied and is probably feeling a positive emotion.
Negative - A negative sentiment means the user who wrote the message seems generally unhappy or dissatisfied and is probably feeling a negative emotion.
Neutral - A neutral sentiment means the user who wrote the message seems to be neither happy nor unhappy and it is difficult to guess their emotion.
Spam Detection Categories:
Spam - Automated spam that is sent to the customer support agents. This topic includes auto-responders, newsletters, guest posts, and other general spam messages that could be ignored by the CS analyst.
Legit - Legitimate conversations in which the user has an actual issue that should be handled by a customer support analyst.
Topic Categories:
Projects - Projects are a collection of related tasks and activities aimed at achieving a specific objective or deliverable which can involve teammate collaboration, time tracking, milestones or goals, and status.
Billing - Billing encompasses managing subscription plans, invoices, payment methods, discounts, plan features, trials, account restrictions, refunds, and more for a seamless billing experience.
Account Management - Account Management covers discussions related to user accounts, including account creation, deletion, updating personal and payment information, and more.
Download the templates at the bottom of this article for AI Categories you can customize and upload as CSV files👇
FAQs
Can I remove an AI category in a workflow?
Can I remove an AI category in a workflow?
No, you can manually apply a new AI category using the “Set AI Detection” action in the workflow builder to override the previous category.
Can I manually change an AI category in the inbox?
Can I manually change an AI category in the inbox?
Yes, you can manually change an AI category from the conversation by selecting the “clear” option for your attribute in the inbox.
Can I use both automatic and manual AI category detection methods in my workflow?
Can I use both automatic and manual AI category detection methods in my workflow?
Yes, you can use both manual and automatic AI category detection methods in your workflow. However, the most recently applied category will take precedence in the conversation.
What happens if the conversation doesn't match any AI category?
What happens if the conversation doesn't match any AI category?
If Fin can’t match a conversation to an AI category, the conversation data attribute won't be set (it's left empty). You can still route these conversations using the "[AI category] is Unknown" predicate in a workflow branch. For clearer reporting, consider creating an “Other” AI category as an explicit fallback when no specific category applies.
AI Category Templates
Need more help? Get support from our Community Forum
Find answers and get help from Intercom Support and Community Experts