Deliver Data Step

The Deliver Data step lets you write data collected during a workflow back to one or more of your organization's datasets. For each dataset, you define which entity to write to, what operation to perform, and how workflow fields map to dataset properties.

You can also write one record per item in a collection (see Repeating Records), and attach files to the records you write (see File-to-record mappings).

Datasets

A Deliver Data step contains one or more dataset operations.

name	type	required	description
`datasetId`	string	yes	The `id` of the target dataset, as returned by the Data Set Management endpoints.
`selectedSearchDataStep`	string or null	yes	The name of the Data Search step whose record you are updating. Use `null` when the step does not update a searched record (for example, a plain `INSERT`). This links the two steps, but it does not identify the record on its own — see Key fields.
`stepDatasetInstanceId`	string or null	yes	A stable unique identifier you assign to this dataset entry within the step. It is used internally to link field mappings to their dataset. Any non-empty string works; the Streamline UI generates a short random value (e.g. `"fg4z6aiy"`). Must not change once field mappings have been created against it.
`entities`	array	yes	One or more entity operations. See Entities below.

Entities

Each entity operation targets a specific entity type within the dataset.

name	type	required	description
`linkId`	string or null	yes	The `linkId` of a traversal route entry in the dataset, used when writing to a linked (non-primary) entity. Corresponds to the same `linkId` used as a key in the Data Search step's `linkedEntitiesFilters`. Use `null` for the primary entity.
`entityId`	string	yes	Identifier of the entity type to write to.
`operation`	string	yes	Operation to perform. See Operations below.
`mappings`	array	yes	Field mappings from workflow data to entity properties. See Mappings below.
`repeatableConfigs`	array or null	no	Collections to write as one record per item. See Repeating Records below.
`fileToEntityMappings`	array or null	no	Files to create and link to the record. See File-to-record mappings below.

Operations

INSERT: Create a new record in the dataset.
UPDATE: Modify an existing record. Requires at least one key field.
UPSERT: Insert if no matching record exists; update if one does.
SKIP: Perform no operation—useful in conditional branches.

Mappings

Each mapping connects a workflow value to a field in the dataset entity.

name	type	required	constraints	description
`fieldId`	string or null	yes		The `id` of the dataset field to write to, as returned by the Data Catalog field endpoints.
`value`	string, number, object, or null	yes		The value to write. Strings support mapping placeholders (e.g. `{{`Form`.`email`}}`). See Date and time values for the object form.
`isKeyField`	boolean	yes		Whether this mapping identifies the record to update. See Key fields. Use `false` for every other mapping.
`primitiveType`	string or null	yes	See Value types	The type of the source value. Must match the type of the mapped workflow field.
`complexType`	string or null	yes	See Value types	The type of the target dataset field.
`repeatableId`	string or null	no	Must match a `repeatableId` in the entity's `repeatableConfigs`	Marks this mapping as part of a repeating collection. See Repeating Records.

Key fields

A key field tells Streamline which existing record to change. Set isKeyField to true on the mapping that carries the record's unique identifier, and false on every other mapping.

The most common pattern is updating a record found by a Data Search step:

Set selectedSearchDataStep to the name of the Data Search step.
Add a mapping whose fieldId is the entity's primary key field (for Salesforce, the Record Id) and whose value maps that same field from the Data Search step.
Set isKeyField to true on that mapping.

📘
When you set up this step in the Streamline app, the key field mapping is added for you. When you configure the step through the API, you must add it yourself. An UPDATE with no key field fails the session with the error "UPDATE operations require at least one usable key field".

Value types

primitiveType accepts:

string, number, boolean, date, file
array<string>, array<number>, array<boolean>, array<date>, array<file>

complexType accepts:

string, object, number, float, binary, boolean
date, time, datetime, uuid, unknown, file
array<string> — use this for multi-select (ordered list) dataset fields.

Both keys are required, but either can be null when the type is not known.

Date and time values

A static date-time can be sent as an object instead of a string:

name	type	required	constraints	description
`utc`	string	yes	ISO 8601 date-time ending in `Z`	The moment in UTC, e.g. `"2026-03-15T14:30:00Z"`.
`timezone`	string	no		IANA time zone name, e.g. `"America/New_York"`.
`offset`	string	no	`+HH:MM` or `-HH:MM`	UTC offset, e.g. `"-04:00"`.

At least one of timezone or offset is required.

Repeating Records

Use Repeating Records to write one record for each item in a collection — for example, one Contact per entry in a webhook's contacts array, or one record per result of a Data Search step with Repeating Records enabled.

Each entry in repeatableConfigs names a collection:

name	type	required	description
`repeatableId`	string	yes	A unique identifier you assign to this collection. Mappings reference it through their `repeatableId`.
`collectionSourcePath`	string	yes	A mapping placeholder pointing to the collection in an earlier step, e.g. {{`Incoming webhook`.`body.contacts`}}.

Mappings that belong to the collection set repeatableId and use [] in their placeholder to refer to the current item, e.g. {{`Incoming webhook`.`body.contacts[].email`}}. Streamline resolves the placeholder once per item and writes one record per item.

A few rules apply:

Every repeatableId used on a mapping must match an entry in the same entity's repeatableConfigs.
Mappings without a repeatableId write the same value to every record.

File-to-record mappings

Use fileToEntityMappings to create file records and link them to the record you write — for example, to attach an uploaded document to a Case.

name	type	required	description
`fileRowId`	string	yes	A unique identifier you assign to this file mapping.
`fileEntityId`	string	yes	The `id` of the file entity to create records in.
`fileLinkId`	string	yes	The `linkId` of the dataset relationship that connects the record to its files.
`fileFieldMappings`	array	no	Mappings for the file entity's fields. Each item has the same structure as a mapping. Default: `[]`.

Examples

Let's start with the simplest valid configuration: inserting a single record with one field.

{
  "datasets": [
    {
      "datasetId": "ds1b2c3d-e5f6-7890-1234-567890abcdef",
      "selectedSearchDataStep": null,
      "stepDatasetInstanceId": "ab1c2d3e",
      "entities": [
        {
          "linkId": null,
          "entityId": "e1b2c3d4-e5f6-7890-1234-567890abcdef",
          "operation": "INSERT",
          "mappings": [
            {
              "fieldId": "f1b2c3d4-e5f6-7890-1234-567890abcdef",
              "value": "[email protected]",
              "isKeyField": false,
              "primitiveType": "string",
              "complexType": "string"
            }
          ]
        }
      ]
    }
  ]
}

The next example updates a customer found by a Data Search step named Search Customers, and inserts an audit log entry in the same step. The first mapping is the key field: it maps the customer's Record Id from the Data Search step to the entity's primary key field.

{
  "datasets": [
    {
      "datasetId": "ds2b3c4d-e5f6-7890-1234-567890abcdef",
      "selectedSearchDataStep": "Search Customers",
      "stepDatasetInstanceId": "ab1c2d3e",
      "entities": [
        {
          "linkId": null,
          "entityId": "e2b3c4d5-e5f6-7890-1234-567890abcdef",
          "operation": "UPDATE",
          "mappings": [
            {
              "fieldId": "f2b3c4d5-e5f6-7890-1234-567890abcdef",
              "value": "{{`Search Customers`.`f2b3c4d5-e5f6-7890-1234-567890abcdef`}}",
              "isKeyField": true,
              "primitiveType": "string",
              "complexType": "string"
            },
            {
              "fieldId": "f3c4d5e6-e5f6-7890-1234-567890abcdef",
              "value": "{{`Form`.`first_name`}} {{`Form`.`last_name`}}",
              "isKeyField": false,
              "primitiveType": "string",
              "complexType": "string"
            },
            {
              "fieldId": "f4d5e6f7-e5f6-7890-1234-567890abcdef",
              "value": "{{`Form`.`email`}}",
              "isKeyField": false,
              "primitiveType": "string",
              "complexType": "string"
            },
            {
              "fieldId": "f5e6f7a8-e5f6-7890-1234-567890abcdef",
              "value": "{{`Form`.`seats`}}",
              "isKeyField": false,
              "primitiveType": "number",
              "complexType": "number"
            },
            {
              "fieldId": "f6f7a8b9-e5f6-7890-1234-567890abcdef",
              "value": "true",
              "isKeyField": false,
              "primitiveType": "boolean",
              "complexType": "boolean"
            }
          ]
        },
        {
          "linkId": null,
          "entityId": "e3c4d5e6-e5f6-7890-1234-567890abcdef",
          "operation": "INSERT",
          "mappings": [
            {
              "fieldId": "f7a8b9c0-e5f6-7890-1234-567890abcdef",
              "value": "{{`Session metadata`.`__session_start_date`}}",
              "isKeyField": false,
              "primitiveType": "string",
              "complexType": "datetime"
            },
            {
              "fieldId": "f8b9c0d1-e5f6-7890-1234-567890abcdef",
              "value": "customer_updated",
              "isKeyField": false,
              "primitiveType": "string",
              "complexType": "string"
            },
            {
              "fieldId": "f9c0d1e2-e5f6-7890-1234-567890abcdef",
              "value": {
                "utc": "2026-03-15T14:30:00Z",
                "timezone": "America/New_York"
              },
              "isKeyField": false,
              "primitiveType": "date",
              "complexType": "datetime"
            }
          ]
        }
      ]
    }
  ]
}

The next example inserts one Contact for each entry in the contacts array sent to an Incoming Webhook step named Incoming webhook. Every Contact also receives the same static source value.

{
  "datasets": [
    {
      "datasetId": "ds3c4d5e-e5f6-7890-1234-567890abcdef",
      "selectedSearchDataStep": null,
      "stepDatasetInstanceId": "cd3e4f5a",
      "entities": [
        {
          "linkId": null,
          "entityId": "e4d5e6f7-e5f6-7890-1234-567890abcdef",
          "operation": "INSERT",
          "repeatableConfigs": [
            {
              "repeatableId": "contacts",
              "collectionSourcePath": "{{`Incoming webhook`.`body.contacts`}}"
            }
          ],
          "mappings": [
            {
              "fieldId": "fa1b2c3d-e5f6-7890-1234-567890abcdef",
              "value": "{{`Incoming webhook`.`body.contacts[].firstName`}}",
              "isKeyField": false,
              "primitiveType": "string",
              "complexType": "string",
              "repeatableId": "contacts"
            },
            {
              "fieldId": "fb2c3d4e-e5f6-7890-1234-567890abcdef",
              "value": "{{`Incoming webhook`.`body.contacts[].email`}}",
              "isKeyField": false,
              "primitiveType": "string",
              "complexType": "string",
              "repeatableId": "contacts"
            },
            {
              "fieldId": "fc3d4e5f-e5f6-7890-1234-567890abcdef",
              "value": "partner_import",
              "isKeyField": false,
              "primitiveType": "string",
              "complexType": "string"
            }
          ]
        }
      ]
    }
  ]
}

The last example inserts a Case and attaches the file uploaded in a form step named Form to it.

{
  "datasets": [
    {
      "datasetId": "ds4d5e6f-e5f6-7890-1234-567890abcdef",
      "selectedSearchDataStep": null,
      "stepDatasetInstanceId": "ef5a6b7c",
      "entities": [
        {
          "linkId": null,
          "entityId": "e5e6f7a8-e5f6-7890-1234-567890abcdef",
          "operation": "INSERT",
          "mappings": [
            {
              "fieldId": "fd4e5f6a-e5f6-7890-1234-567890abcdef",
              "value": "{{`Form`.`subject`}}",
              "isKeyField": false,
              "primitiveType": "string",
              "complexType": "string"
            }
          ],
          "fileToEntityMappings": [
            {
              "fileRowId": "case-attachment",
              "fileEntityId": "e6f7a8b9-e5f6-7890-1234-567890abcdef",
              "fileLinkId": "l6f7a8b9-e5f6-7890-1234-567890abcdef",
              "fileFieldMappings": [
                {
                  "fieldId": "fe5f6a7b-e5f6-7890-1234-567890abcdef",
                  "value": "{{`Form`.`attachment`}}",
                  "isKeyField": false,
                  "primitiveType": "file",
                  "complexType": "file"
                }
              ]
            }
          ]
        }
      ]
    }
  ]
}