Identity graph

An identity graph is the project-level configuration behind a Profile Resolution project.

It defines the source, profile models, identifiers, matching rules, golden fields, schedule, and output tables used to create the resolved profile spine.

In the DinMo UI, users do not need a separate workflow to create an identity graph. Creating or editing a Profile Resolution project creates and updates the identity graph.

What an identity graph contains

An identity graph contains:

• one source where Identity Resolution runs

• the profile-like models included in the project

• the identifiers used to connect records

• standardization methods and conflict limits

• matching rules and criteria

• golden fields and survivorship strategies

• the schedule used to refresh the output

When to create more than one

Use one identity graph when a single matching strategy can resolve the same entity type.

Create separate identity graphs when you need materially different resolution logic, such as:

• people vs companies

• customers vs households

• regional datasets with different identifiers

• test projects vs production projects

Where to configure it

Use Create your first Profile Resolution project for the setup workflow.

After the first run, use Review and monitor to validate quality, inspect runs, and decide whether the resolved profile spine is ready for downstream use.

For the rule model, see Matching rules. For generated warehouse outputs, see Output tables.