Conversations

A conversation is a thread of question-answer interactions between a user and the AI assistant. Kapa persists all conversations across your deployments so you can review and export them. To make working with your conversation data easier, each conversation is automatically enriched:

• Intent tags: Kapa automatically assigns Intent Tags to identify what kind of conversation it is — for example, whether a user is troubleshooting an issue, discovering how to use your product, or asking about a competitor.
• Custom tags: Define your own tags to classify conversations into categories that matter to your business — for example, by product area, feature, or team. See Custom Tags for details.

Intent Tags

Kapa automatically assigns tags to conversations by analyzing the intent of the questions in the thread.

The available Intent Tag categories are:

Category	Description
Uncertain	Conversations where Kapa expresses low confidence or cannot find sufficient information in your knowledge sources to provide a complete answer.
Troubleshooting	Conversations where users describe errors, unexpected behavior, or implementation issues they're experiencing with your product.
Unsupported Feature	Conversations where users inquire about functionality that isn't currently available in your product offering.
Competitor	Conversations that mention or compare your product with competing solutions in the market.
Off Topic	Conversations unrelated to your product, documentation, or support scope.
Deflection Success	Conversations where a user initially intended to submit a support ticket, but the Support Form Deflector provided a satisfactory answer, eliminating the need for human intervention.
Deflection Failure	Conversations where a user proceeded to submit a support ticket after receiving an answer from the Support Form Deflector, indicating the automated response didn't resolve their issue.

Custom Tags

In addition to the built-in Intent Tags, you can define your own custom tags to classify conversations into categories specific to your business. Common use cases include:

• Product tags — classify by product when your company offers multiple distinct products or product areas (e.g. Container Registry, Object Storage, Managed Database).
• Feature tags — track specific features or capabilities within a product (e.g. replication policies, automated backups, connection pooling).
• Administrative tags — track conversations about account management, pricing, compliance, or support (e.g. Billing & Invoices, Compliance & Privacy).

Custom tags are often used to route relevant questions to the product managers or teams responsible for that area of the product.

Each custom tag has a name and a description that tells the AI when the tag should be applied. A single conversation can have multiple custom tags assigned, or none if the conversation doesn't match any of your tag descriptions. You can define up to 20 custom tags per project.

Custom tags can be managed from the Manage Tags page in the Kapa platform.

Setting up custom tags for the first time

When you define custom tags for the first time, Kapa automatically applies them to conversations from the last 7 days. This backfill happens within roughly 30 minutes, so you can see results on recent conversations right away.

Managing custom tags

Whenever you create or edit any custom tag, Kapa waits 10 minutes after your last change before applying tags, so you can take your time refining descriptions without triggering intermediate runs. Tags are applied in batches roughly every 30 minutes, so expect new conversations to be labeled within about an hour of being created. After the initial backfill, new or edited tags are only applied to conversations going forward.

When you edit a custom tag, the change applies to all historical assignments as well. Only edit a tag if you are not changing its meaning (e.g. fixing a typo). If you want to change the meaning of a tag, create a new tag and delete the old one instead.

When you delete a custom tag, it is soft-deleted so that historical assignments remain visible on past conversations. The tag is no longer applied to new conversations.

The quality of your tag descriptions directly impacts classification accuracy. See the guide on how to write good custom tags.

Conversations

The conversations page contains all raw conversations across all deployment integrations for your project. By default, it lists the conversations that occurred in the past 30 days in reverse chronological order.

Filters

The conversation list can be filtered using several criteria to help you find specific conversations:

• Text contents: Search for specific words or phrases in questions and answers
• Integration: Filter by specific Kapa deployment (e.g., Website Widget, Slack bot)
• Uncertainty: Show only uncertain or certain conversations
• Intent Tags: Filter by automatically assigned intent categories
• Custom Tags: Filter by your own custom tag categories
• Feedback: Filter by upvotes, downvotes, or comments
• Status Tags: Find conversations with specific status indicators
• Deflection Status: Filter by deflection outcome (only populated for conversations from the Support Form Deflector)
• Date: Limit results to a specific time period

Each Kapa deployment must use a unique integration ID for proper analytics tracking. If multiple deployments (such as your documentation website and marketing website) share the same integration ID for their Website Widgets, you won't be able to distinguish between them in the analytics.

tip

If you see conversations labelled with "Unknown" integration, it means the integration_id was not included in the API request. To fix this, pass a valid integration_id with every API call. Managed integrations like the Website Widget and Slack bot handle this automatically.

Export conversations

You can export conversation data to a CSV file for further analysis or reporting purposes.

To export conversation data to a CSV file:

1. Set the desired date range using the date picker
2. Apply any filters to narrow down the data set
3. Click the Export to CSV button next to the date picker

The exported CSV contains all conversations matching your current filters and date range. Each row in the CSV represents a single Q/A interaction with the following columns:

Column	Description
Thread ID	Unique identifier for the conversation thread
Question/Answer ID	Unique identifier for the individual Q/A interaction
Timestamp (UTC)	When the question was asked
Question	The user's question text
Answer	Kapa's response text
Query Type	The type of query
Language	Detected language of the question
Is Uncertain	Whether Kapa expressed uncertainty in the answer
Intent Tags	Comma-separated list of Intent Tags
Custom Tags	Comma-separated list of custom tags assigned to the conversation
Answer Upvotes	Number of upvote reactions
Answer Downvotes	Number of downvote reactions
Feedback Comments	User-provided text feedback comments
CSAT Rating	Customer satisfaction rating (if collected)
CSAT Comment	Customer satisfaction comment text (if collected)
Support Form Deflection Status	Success/Failure status for Support Form Deflector questions
Integration	Name of the integration where the question was asked
Status	Current status tag for the conversation
Question Origin URL	URL of the page where the question was asked
End User ID	Unique identifier for the end user
End User Email	End user's email address (if available)
End User Unique Client ID	Custom client identifier for the end user (if set)
File Attachments	Comma-separated list of attached file names
URL	Direct link to view the conversation in the Kapa platform