Unify and manage your data

AI automapping API

Automatically generate a mapping from a source file using AI-powered assistance.

The AI automapping API helps you automate the process of mapping source file columns to Reltio entity type attributes. Instead of manually selecting attribute mappings during data loading, use this API to receive intelligent suggestions based on column headers and a few sample data rows.

This API leverages language models to analyze column names and sample values. It returns structured mapping recommendations for each column, along with a confidence level and reasoning. It returns only high or medium confidence mappings to help reduce errors and ensure quality. Suggestions are not generated for columns with low or ambiguous match confidence.

Use the AI automapping API when working with large or unfamiliar data files. It accelerates the setup process and improves consistency by automatically suggesting how your data aligns with your tenant's L3 configuration.

Parameters

Operation endpoint:

POST {dl_host}/api/ai/{tenant}/map
CategoryParameterRequiredDescription
HeadersAuthorizationYesBearer token for authentication. Format: Bearer <accessToken>.
Content-TypeYesMust be application/json.
EnvironmentURLYesThe Reltio environment URL where the request is executed.
PathtenantYesYour tenant name in the Reltio environment.
BodycolumnsYesComma-separated string of source column headers.
dataYesArray of sample data rows for context.
typeYesURI of the entity type, for example configuration/entityTypes/HCP.

Request

{
  "columns": "First Name,Last Name,Email,Phone",
  "data": [
    "John,Doe,john.doe@example.com,1234567890",
    "Jane,Smith,jane.smith@example.com,9876543210"
  ],
  "type": "configuration/entityTypes/HCP"
}

Response

The API returns a JSON object containing:

  • sources: Original source column definitions with generated keys.
  • attributes: Suggested attribute mappings, each with path, confidence, reasoning, and mapping expression.
  • objectType: The entity type used in the mapping.

Only columns with high or medium confidence matches are included in the response.

{
  "sources": [
    { "name": "First Name", "value": "=getCols(0)", "key": "col.0" },
    { "name": "Last Name", "value": "=getCols(1)", "key": "col.1" },
    { "name": "Email", "value": "=getCols(2)", "key": "col.2" },
    { "name": "Phone", "value": "=getCols(3)", "key": "col.3" }
  ],
  "attributes": [
    {
      "path": "attributes.FirstName",
      "confidence": "high",
      "reasoning": "Exact match for first name attribute",
      "value": "=source('col.0')"
    },
    {
      "path": "attributes.LastName",
      "confidence": "high",
      "reasoning": "Exact match for last name attribute",
      "value": "=source('col.1')"
    },
    {
      "path": "attributes.Email[0].Email",
      "confidence": "high",
      "reasoning": "Email address maps to Email nested attribute for email value",
      "value": "=source('col.2')"
    },
    {
      "path": "attributes.Phone[0].Number",
      "confidence": "high",
      "reasoning": "Phone number maps to Phone nested attribute for phone number value",
      "value": "=source('col.3')"
    }
  ],
  "objectType": "configuration/entityTypes/HCP"
}

}

Availability

  • The AI automapping API is available only if your tenant configuration includes:
  • agentsConfig.aiEnabled: true