For the complete documentation index, see llms.txt. This page is also available as Markdown.

Confluence Configuration

Confluence integration required app registration in Developer console

This is the Atlassian dashboard menu:

Authorization

Navigate to Authorization and enable OAuth 2.0 (3LO).

Configure the following:

  • Callback URL – The URL where Atlassian redirects users after authorization.

    • Example: https://kbot-preprod.concentrix.com/connectors/oauth/callback

Permissions

Navigate to Permissions → Confluence API.

Classic Scopes

Classic scopes provide broad access to Confluence resources and are the recommended option for most integrations because they are simpler to configure and maintain.

Required classic scopes:

Granular Scopes

Granular scopes allow you to grant access to individual operations instead of broad capabilities. They follow the principle of least privilege but require selecting and maintaining a larger number of permissions.

Required Granular scopes:

Scope Name

Code

View pages

read:page:confluence

View folders

read:folder:confluence

View and download content attachments

read:attachment:confluence

View content summaries

read:content.metadata:confluence

View spaces

read:space:confluence

Callback:

US instance: https://us.knowledge.ixhello.com/connectors/oauth/callback

EU instance:

https://eu.knowledge.ixhello.com/connectors/oauth/callback

Distribution

You need to select to be Sharing and also add the vendor name and Privacy policy

Credentials

After saving the configuration, copy the following values from the Settings page:

  • Client ID

  • Client Secret

These credentials are required by the application to complete the OAuth 2.0 authorization flow and exchange authorization codes for access tokens.

Confluence Knowledge Bot Reporting and Analytics

The Reporting section provides insights into Knowledge Bot (K-Bot) usage, user interactions, feedback, token consumption, and operational costs. This information helps administrators monitor bot performance and evaluate the quality of generated responses.

Access Control: Reporting data is available only to users with Admin or Super Admin roles. Access is governed through role-based permissions.

Configure a Confluence Data Source

To connect a Confluence instance as a knowledge source:

  1. Navigate to the Data Source Configuration page.

  2. Enter the required Client ID and Client Secret credentials.

  3. Save the configuration.

These credentials allow the system to securely connect to Confluence and retrieve content for ingestion.

Viewing Ingested Knowledge Data

  1. Save the configuration.

  2. The system automatically retrieves and processes the content.

  3. Once processing is complete, the ingested content is divided into searchable chunks and made available to the Knowledge Bot.

Ingest Content from a Confluence Page

After configuring the Confluence data source:

  1. Go to the Upload section.

  2. Enter the URL of the Confluence page you want to ingest.

  3. Save the configuration.

The system will begin processing the page and extracting its content.

Processing Results

Once ingestion is complete, you can:

  • View the ingested content chunks.

  • Verify that the page content has been successfully processed.

  • Confirm that the data is available for use by the Knowledge Bot.

Querying the Knowledge Bot

After the content has been ingested:

  1. Open the Chatbot page.

  2. Select the Knowledge Bot associated with the uploaded data.

  3. Ask questions related to the ingested documentation.

Source References

Each response includes:

  • The generated answer.

  • A reference to the source document used to generate the response.

This helps users verify where the information originated.

Providing Feedback

Users can provide feedback on chatbot responses directly from the chat interface.

Feedback helps evaluate response quality and improve the effectiveness of the knowledge base.'

Reporting

The Reporting section provides analytics and usage insights for Knowledge Bots.

Available Metrics

Reporting includes:

  • Total requests

  • Positive and negative feedback

  • Total token usage

  • Model costs

  • Request statistics

  • Other usage-related metrics

Users can review detailed information for each interaction, including:

  • User question

  • Generated response

  • Rephrased question

  • Model usage information

Question Rephrasing Details

For each query, the system may generate a rephrased version of the user's question to improve retrieval quality.

Reporting displays:

  • Original question

  • Rephrased question

  • Embedding model used (for example, Text Embedding 3 Large)

  • Associated processing details


Chat Logs

The Chat Logs section provides a detailed history of conversations.

For each conversation, administrators can view:

  • User questions

  • Bot responses

  • Response formatting

  • Conversation details

  • Associated metadata

This enables troubleshooting, auditing, and review of chatbot interactions.


Exporting Chat Logs

Chat logs can be exported for external analysis and record keeping.

Export Process

  1. Open the Chat Logs section.

  2. Select Export.

  3. The export is processed asynchronously.

  4. Once processing is complete, the file becomes available for download.

Exported Data

The exported file may include:

  • Feedback details

  • Bot responses

  • Original questions

  • Rephrased questions

  • Language information

  • Timestamps

  • Knowledge Bot name

  • Chat Log ID

  • Other conversation metadata


Access Control

Role-Based Access

Access to reporting and analytics features is controlled through role-based permissions.

Users with Access

The following roles can access the Reporting and Chat Log sections:

Role
Reporting Access

User

No

Admin

Yes

Super Admin

Yes

Note

Super Admins have access to all platform features by default. Since application-level Admins can also access Reporting and Chat Logs, this access behavior should be documented as part of the platform's role-based access model.

Last updated

Was this helpful?