🔍 Understanding Connections, Data Sources, and Partitions in Worknet

Worknet enables powerful AI-driven search and automation by integrating with a wide range of systems. To configure your environment effectively, it's important to understand three foundational concepts: Connections, Data Sources, and Partitions.

🔗 Connection

A Connection in Worknet refers to an authenticated link between Worknet and a backend system. It enables Worknet to securely access and retrieve data from external platforms.

Key Features:

• Authentication Maintained: Connections manage authentication tokens, ensuring continued access without manual intervention.
• Out-of-the-box Integrations: Worknet currently supports native integrations with:
  • Zendesk
  • Salesforce
  • HubSpot
  • Slack
• Token-based Authentication Support: Beyond native integrations, Worknet can easily connect to any system that uses token-based authentication.

Example: If your organization uses a custom CRM with OAuth tokens, you can configure a Connection to that system just as easily as you would with Salesforce.

📡 Data Source

A Data Source defines the actual data that Worknet extracts for AI processing. This data can come from various origins and may or may not require authentication.

Types of Data Sources:

1. Connection-Based Sources: Data pulled from authenticated systems (e.g., Zendesk tickets, Salesforce records).
2. Public Sources: Publicly accessible information (e.g., help center articles or documentation websites).
3. Static Files: Uploaded documents such as PDFs, CSVs, or text files.

Airbyte Integration:

Worknet supports Airbyte as a data ingestion layer. This allows easy connection to any Airbyte-compatible data source.

• Examples: Jira and Confluence integrations in Worknet are powered by Airbyte.
• Flexibility: If Airbyte supports it, you can use it as a Worknet Data Source.

Customization:

Each Data Source can be configured with extraction rules, such as:

• Time-based limits (e.g., pull Zendesk tickets from the past 90 days)
• Field selection (e.g., only pull closed tickets or specific custom fields)

⚠️ Note: Some advanced settings related to data extraction—such as field mapping, filtering logic, or historical depth—may only be configurable by your Customer Success Manager (CSM). Please reach out to your CSM for help with these customizations.

🗂 Partition

A Partition is a logical grouping of Data Sources. When AI actions in Worknet perform a search, they do so within one or more defined Partitions.

Why Use Partitions?

• Controlled Access: Define what data is available for different AI actions or user roles.
• Data Segmentation: Keep internal and external content separated for compliance or UX reasons.

Common Use Cases:

• External-facing Partition: Includes only public or curated content (e.g., Help Center articles) used for customer-facing AI bots.
• Internal-facing Partition: Adds private data sources such as Jira, Confluence, or Slack messages for internal team use.

Example: A customer support AI might only query an external-facing partition, while a technical success engineer’s AI assistant has access to internal-facing partitions as well.

✅ Summary

💬 Need Help?

For help setting up Connections, Data Sources, or Partitions in your Worknet environment, contact your Customer Success Manager or email us at support@worknet.ai.

We’re here to help you get the most out of Worknet.