Skip to main content
Bonds define how data in one system relates to data in another. They’re the foundation for cross-system workflows in BonData.

​
Bonds vs. Suggested Bonds

Description
BondsConfirmed connections between systems. Specify which fields are linked and how records match.
Suggested BondsAI-generated recommendations. BonData analyzes metadata and hashed data to find potential relationships, each with a confidence score.

​
The Bonds page

Navigate to Bonds in the sidebar to manage all your data relationships.

​
Active tab

Shows all confirmed bonds with:
  • Source and target integrations and fields
  • The Default field used for matching
  • Bond Type — probability-based or compounded

​
Suggested tab

Shows AI-discovered connections with:
  • Matching Values — confidence score
  • Actions — approve or reject via the three-dots menu

​
Create a Bond manually

You can define custom relationships between any two systems.
1

Click Add Bond

On the Bonds page, click Add Bond (top right).
2

Select systems and fields

Pick two systems, then choose which fields to link. Add multiple field pairs for complex matching (e.g., match only when both First Name and Last Name align).
3

Set matching rules

Choose exact match or partial match. For partial matches, set a confidence threshold (e.g., 90–100%) to handle typos and minor differences.
Manual bonds work exactly like confirmed bonds — they appear in the Active tab and are available for use in Agents.

​
Data Map

The Data Map (Data > Data Map) provides a visual representation of all your bonds.
Click any entity to open a drawer with its active and suggested bonds. You can approve suggestions directly from here.

​
Filtering

Use the search and filter panel on the left to narrow down by system, entity, or bond type.

​
Data security

BonData does not store raw data — only metadata, connection details, and hashed values. Data is accessed only when executing Agents. The platform is GDPR and SOC II compliant.