Skip to content

Domain ontology

An ontology gives source records a common domain vocabulary. A team can define a Prospect entity, its identifier and lifecycle stage, and its relationship to an organization. Agents and apps can then interpret records using the same terms.

Inventory the organization’s apps with a specialist. For each source, identify the record type, stable ID, relevant dates, and fields that are safe to use. Choose domain attribute names that survive a change of provider. Use a required identity field when the team needs to reconcile the same record over repeated reads. Attribute types and required fields are part of the model the mappings should satisfy.

Define relationships by naming both entity types, the direction, and the cardinality where known. Review the rule that joins records before using it to publish or mutate data. For a recurring calendar event, two occurrences can share a title. Use an occurrence identifier when mapping results into Meeting and ask for specialist review when the source has conflicting identities.

This is the Meeting desk ontology fixture. A connector may use different field names. The mapping belongs with the feed or playbook that reads it.

{
"schemaVersion": 1,
"resourceKey": "ontology.persona.meeting-desk",
"version": "1.0.0",
"name": "Meeting desk",
"status": "defined",
"entities": [
{
"type": "Meeting",
"description": "An upcoming occurrence on an authorized calendar.",
"attributes": {
"occurrenceId": {
"type": "string",
"required": true
},
"title": {
"type": "string"
},
"startsAt": {
"type": "datetime",
"required": true
}
}
}
],
"relationships": []
}
Consumer How it uses the mapping
Custom app Displays normalized dates and titles from permitted records
Agent Reads the same domain fields as playbook input and proposes a change through an allowed operation
Specialist Reviews missing or conflicting identities and validates a revised mapping

See feed definitions and records and apps.