Industry & Department Solutions - RPA integration contracts

industry-department-solutions

latest

false

Financial Services Solutions user guide

RPA integration contracts

Data contracts for customer-built RPAs that integrate with the Loan Setup capability, defining the input and output schema each RPA must satisfy to extract and deliver data from your systems of record.

This page defines the data contracts for customer-built RPAs that integrate with the Loan Setup capability. Customers are responsible for building RPAs that extract data from their systems of record and deliver it in the formats specified below.

Overview

The Loan Setup capability requires data from multiple banking systems that are unique to your institution. To connect it to your systems, RPAs must be built that:

Call the Maestro process API endpoint with a loan_id to start processing
Call the Maestro process Mark Loan as Completed API endpoint to signal that a loan phase is finished
Pull document data from the systems of record and write to Orchestrator storage buckets
Pull JSON/XML data from the systems of record
Transform data into the standardized formats defined in this document

RPA registry pattern

Customers should register their data integration RPAs with the solution by placing their RPA type, path, and related details into the Data Fabric FINSProcessRegistry table. The Maestro process will invoke these RPAs by name during execution.

FINSProcessRegistry table schema

Field	Type	Required	Unique	Description
`Id`	string	Yes	No	Internal primary key
`RPAName`	NVARCHAR(200)	Yes	Yes	RPA identifier used for invocation
`DisplayName`	NVARCHAR(500)	No	No	Human-readable name
`Description`	NVARCHAR(2000)	No	No	What this RPA does
`OrchestratorFolderPath`	NVARCHAR(500)	Yes	No	Orchestrator folder path where the process is deployed
`ProcessName`	NVARCHAR(200)	Yes	No	Process name in Orchestrator
`ProcessType`	FINSProcessType	Yes	No	Type: `DocumentRetrieval`, `DataRetrieval`, or `DataParse`
`FindingCategory`	choice set	Yes	No	Finding category, such as `Extraction`, `LoanOfficerEmail`, `ManagerEmail`, `Summary`, or `Risk`
`RPATypeVersion`	NVARCHAR(50)	Yes	No	Version of the RPA type contract (for example, `1`)
`EntryPointName`	NVARCHAR(200)	Yes	No	Entry point name within the process
`ProcessInputTemplate`	NVARCHAR	No	No	JSON template with variables that Maestro fills when it invokes the process
`IsActive`	BIT	Yes	No	Whether the RPA is enabled
`CreatedAt`	DATETIMEOFFSET	Yes	No	When the RPA was registered

RPA types

The customer-built data integration RPAs use the following three standard ProcessType values.

RPAType	Purpose	When invoked
DataRetrieval	Pull structured data from core banking, LOS, or other systems	After loan trigger
DataParse	Transform or parse retrieved data into standardized formats, once per source	After DataRetrieval
DocumentRetrieval	Pull the loan's document files and write them to the storage bucket under the provided prefix	After loan trigger

Process Trigger Contract

To start the Loan Setup workflow, the customer's dispatcher RPA calls the Maestro process's HTTP API endpoint when a new loan requires processing. UiPath does not provide this dispatcher; your institution builds and maintains it. UiPath provides the exact API endpoint URL after deployment.

Request payload example

{
  "version": "1.0.0",
  "loan_id": "LOAN-2025-001234"
}
{
  "version": "1.0.0",
  "loan_id": "LOAN-2025-001234"
}

Field definitions

Field	Type	Required	Description
`version`	string	Yes	Contract version as a semantic-versioning string (for example, `1.0.0`).
`loan_id`	string	Yes	Unique loan identifier from the Loan Origination System. This ID is used to correlate all data extraction and agent outputs.

Implementation notes

Dispatcher responsibility: Your institution builds and maintains the dispatcher RPA or process that monitors the LOS for loan changes and calls the API trigger. UiPath does not provide this dispatcher.
Idempotency: If the same loan_id is submitted multiple times, the solution processes it again. Preventing duplicate submissions is the dispatcher's responsibility.

Authentication and invocation

The trigger is protected by UiPath Identity. The customer-built dispatcher is registered as a confidential external application, and its client secret belongs in a secure, server-side secret store. The secret must never be embedded in a browser application or in source control.

Required registration

The dispatcher requires a Confidential application registered in UiPath Admin with the OR.Jobs.Write scope. Its generated client ID and client secret are the credentials used to obtain an access token, and both are stored securely.

Token request

The client credentials are exchanged for a short-lived access token:

curl --location 'https://cloud.uipath.com/identity_/connect/token' \
  --header 'Content-Type: application/x-www-form-urlencoded' \
  --data-urlencode 'client_id=<external-app-client-id>' \
  --data-urlencode 'client_secret=<external-app-client-secret>' \
  --data-urlencode 'grant_type=client_credentials'
curl --location 'https://cloud.uipath.com/identity_/connect/token' \
  --header 'Content-Type: application/x-www-form-urlencoded' \
  --data-urlencode 'client_id=<external-app-client-id>' \
  --data-urlencode 'client_secret=<external-app-client-secret>' \
  --data-urlencode 'grant_type=client_credentials'

Trigger invocation

The Maestro endpoint that UiPath provides after deployment accepts the access token as a bearer token:

curl --location 'https://cloud.uipath.com/<organization-id>/<tenant-name>/orchestrator_/t/<agentic-process-trigger-id>/agenticprocess' \
  --header 'Content-Type: application/json' \
  --header 'Authorization: Bearer <access-token>' \
  --data '{"version":"1.0.0","loan_id":"LOAN-2025-001234"}'
curl --location 'https://cloud.uipath.com/<organization-id>/<tenant-name>/orchestrator_/t/<agentic-process-trigger-id>/agenticprocess' \
  --header 'Content-Type: application/json' \
  --header 'Authorization: Bearer <access-token>' \
  --data '{"version":"1.0.0","loan_id":"LOAN-2025-001234"}'

Each token is refreshed before the next invocation, or cached only for the lifetime returned in the token response.

Mark Loan as Completed Contract

After a loan has finished a lifecycle phase, the customer's RPA calls a separate API endpoint to signal completion. Loan Setup updates the loan status to Completed. This endpoint is idempotent: if the loan is already in that state, it returns success with already_completed=true and takes no further action.

Request payload example

{
  "loan_id": "LOAN-2025-001234",
  "loan_phase": "Origination",
  "loan_type": "HELOC"
}
{
  "loan_id": "LOAN-2025-001234",
  "loan_phase": "Origination",
  "loan_type": "HELOC"
}

Field definitions

Field	Type	Required	Description
`loan_id`	string	Yes	Loan identifier matching the one used in the Process Trigger.
`loan_phase`	string	Yes	Phase being marked complete, for example `Origination` or `QAQC`. Scopes the completion event to a specific stage of the loan lifecycle.
`loan_type`	string	Yes	Loan product type, for example `HELOC` or `Mortgage`. Recorded in the audit log alongside the completion event.

Implementation notes

Idempotency: If the loan is already in Completed state, the call returns success with already_completed=true and takes no further action.
Dispatcher responsibility: Your institution builds and maintains the process that calls this endpoint. UiPath does not provide this dispatcher.

Data Retrieval RPA Contract

This RPA retrieves raw data from multiple source systems, including Core Banking and the LOS, in a single call and returns it to Loan Setup.

Important:

This RPA returns the raw data exactly as it comes from each source system. The data transformation and mapping happen in the Data Parse RPA Contract.

RPA input example

{
  "version": "1.0.0",
  "loan_id": "LOAN-2025-001234"
}
{
  "version": "1.0.0",
  "loan_id": "LOAN-2025-001234"
}

Input fields

Field	Type	Required	Description
`version`	string	Yes	Contract version as a semantic-versioning string (for example, `1.0.0`)
`loan_id`	string	Yes	Loan identifier to retrieve data for

RPA output example

Returns raw data from each source. The custom_data structure varies by source system.

{
  "version": "1.0.0",
  "loan_id": "LOAN-2025-001234",
  "external_loan_id": "001234-02",
  "success": true,
  "sources": [
    {
      "friendly_name": "los",
      "source_system": "LOS",
      "extraction_timestamp": "2025-01-15T11:00:00Z",
      "success": true,
      "custom_data": {
        "source_loan_id": "001234-02",
        "account_number": "001234",
        "application_id": "02",
        "app_date": "2024-11-05",
        "loan_amount": "42900.00",
        "appraised_amount": "564000.00",
        "ltv_ratio": "31.638%",
        "new_expense_ratio": "290.970%",
        "occupancy_type": "Owner-Occupied",
        "purpose_of_loan": "Home Improvement",
        "loan_type": "HELOC",
        "disclosures_date": "2024-11-04",
        "due_date_for_disclosures": "2024-11-08",
        "loan_officer_name": "John Smith",
        "loan_officer_email": "jsmith@examplecu.com",
        "assigned_user": "Mary Johnson",
        "assigned_user_email": "mjohnson@examplecu.com",
        "qualifying_credit_score": "712",
        "applicants": [{"...": "..."}],
        "credit_reports": [{"...": "..."}],
        "property": {"...": "..."},
        "compliance": {"...": "..."},
        "application_notes": [{"...": "..."}]
      }
    }
  ]
}
{
  "version": "1.0.0",
  "loan_id": "LOAN-2025-001234",
  "external_loan_id": "001234-02",
  "success": true,
  "sources": [
    {
      "friendly_name": "los",
      "source_system": "LOS",
      "extraction_timestamp": "2025-01-15T11:00:00Z",
      "success": true,
      "custom_data": {
        "source_loan_id": "001234-02",
        "account_number": "001234",
        "application_id": "02",
        "app_date": "2024-11-05",
        "loan_amount": "42900.00",
        "appraised_amount": "564000.00",
        "ltv_ratio": "31.638%",
        "new_expense_ratio": "290.970%",
        "occupancy_type": "Owner-Occupied",
        "purpose_of_loan": "Home Improvement",
        "loan_type": "HELOC",
        "disclosures_date": "2024-11-04",
        "due_date_for_disclosures": "2024-11-08",
        "loan_officer_name": "John Smith",
        "loan_officer_email": "jsmith@examplecu.com",
        "assigned_user": "Mary Johnson",
        "assigned_user_email": "mjohnson@examplecu.com",
        "qualifying_credit_score": "712",
        "applicants": [{"...": "..."}],
        "credit_reports": [{"...": "..."}],
        "property": {"...": "..."},
        "compliance": {"...": "..."},
        "application_notes": [{"...": "..."}]
      }
    }
  ]
}

The abbreviated nested values show the shape of custom_data. The complete contents vary by source system.

Output fields

Field	Type	Description
`version`	string	Schema version as a semantic-versioning string
`loan_id`	string	Loan identifier from the LOS
`external_loan_id`	string	Source-system loan identifier
`success`	boolean	Whether retrieval completed successfully
`sources[]`	array	Array of source data objects
`sources[].friendly_name`	string	Logical name for this source (e.g., `core`, `los`)
`sources[].source_system`	string	Actual system name (e.g., `Symitar`, `Encompass`)
`sources[].extraction_timestamp`	datetime	When the data was extracted
`sources[].success`	boolean	Whether this source was queried successfully
`sources[].custom_data`	object	Raw data from source; structure varies by system
`sources[].error_code` / `sources[].error_message`	string	Error details if `success` is `false`

Data Parse RPA Contract

This RPA transforms raw data from a single source system into the well-known schema that Loan Setup expects. Loan Setup calls this RPA once for each source returned by DataRetrieval.

RPA input example

Loan Setup passes one source's raw data at a time.

{
  "version": "1.0.0",
  "loan_id": "LOAN-2025-001234",
  "external_loan_id": "001234-02",
  "friendly_name": "los",
  "source_system": "LOS",
  "extraction_timestamp": "2025-01-15T11:00:00Z",
  "custom_data": {
    "source_loan_id": "001234-02",
    "account_number": "001234",
    "application_id": "02",
    "app_date": "2024-11-05",
    "loan_amount": "42900.00",
    "...": "..."
  }
}
{
  "version": "1.0.0",
  "loan_id": "LOAN-2025-001234",
  "external_loan_id": "001234-02",
  "friendly_name": "los",
  "source_system": "LOS",
  "extraction_timestamp": "2025-01-15T11:00:00Z",
  "custom_data": {
    "source_loan_id": "001234-02",
    "account_number": "001234",
    "application_id": "02",
    "app_date": "2024-11-05",
    "loan_amount": "42900.00",
    "...": "..."
  }
}

Input fields

Field	Type	Description
`version`	string	Contract version as a semantic-versioning string (for example, `1.0.0`)
`loan_id`	string	Loan identifier for correlation
`external_loan_id`	string	Source-system loan identifier used to trace the record back to the LOS
`friendly_name`	string	Logical name for this source (e.g., `core`, `los`)
`source_system`	string	Actual system name (e.g., `Symitar`, `Encompass`)
`extraction_timestamp`	datetime	When the data was extracted
`custom_data`	object	Raw data from the source system to be parsed/mapped

RPA output example

Returns data mapped to the well-known schema fields that Loan Setup consumes.

{
  "success": true,
  "loan_id": "LOAN-2025-001234",
  "external_loan_id": "001234-02",
  "friendly_name": "core",
  "source_system": "LOS",
  "extraction_timestamp": "2025-01-15T11:00:00Z",
  "mapped_data": {
    "source": "LOS",
    "source_loan_id": "001234-02",
    "version": "2.0.0",
    "custom_data": {
      "channel": "branch",
      "branch_id": "TPA-04"
    },
    "loan_info": {
      "loan_type": "HELOC",
      "loan_application_id": "001234-02",
      "account_number": "001234",
      "application_id": "02",
      "app_date": "2024-11-05",
      "loan_amount": "42900.00",
      "appraised_amount": "564000.00",
      "ltv_ratio": "31.638%",
      "new_expense_ratio": "290.970%",
      "occupancy_type": "Primary Residence",
      "purpose_of_loan": "Home Improvement",
      "disclosures_date": "2024-11-04",
      "due_date_for_disclosures": "2024-11-08",
      "loan_officer_name": "John Smith",
      "loan_officer_email": "jsmith@examplecu.com",
      "assigned_user": "Mary Johnson",
      "assigned_user_email": "mjohnson@examplecu.com",
      "qualifying_credit_score": "712",
      "custom_data": {
        "internal_promo_code": "REFI-2024-Q4",
        "underwriter_notes_ref": "UW-7741"
      }
    },
    "applicant": [
      {
        "applicant_type": "Main Applicant",
        "first_name": "JOHN",
        "middle_name": "A",
        "last_name": "DOE",
        "applicant_email": "john.doe@example.com",
        "ssn": "589050676",
        "date_of_birth": "1969-03-27",
        "applicant_street": "16039 PENWOOD DR",
        "applicant_city": "TAMPA",
        "applicant_state": "FL",
        "applicant_zip": "33647-1139",
        "occupancy_status": "Own",
        "employer_name": "GINGER HEALTH",
        "position": "PHYSICAL THERAPY",
        "employment_status": "Employed",
        "years_employed": "22",
        "monthly_income": "3511.11",
        "custom_data": {
          "employer_ein": "12-3456789",
          "los_applicant_id": "APP-1001"
        }
      },
      {
        "applicant_type": "Co-Borrower",
        "first_name": "JANE",
        "middle_name": "B",
        "last_name": "DOE",
        "ssn": "412738955",
        "date_of_birth": "1971-08-14",
        "applicant_street": "742 EVERGREEN TERRACE",
        "applicant_city": "ORLANDO",
        "applicant_state": "FL",
        "applicant_zip": "32801",
        "occupancy_status": "Rent",
        "employer_name": "ORLANDO REGIONAL MEDICAL CENTER",
        "position": "REGISTERED NURSE",
        "employment_status": "Employed",
        "years_employed": "12",
        "monthly_income": "5200.00"
      }
    ],
    "credit_report": [
      {
        "credit_report_first_name": "JOHN",
        "credit_report_last_name": "DOE",
        "credit_report_request_date": "2024-10-15",
        "credit_score": "750",
        "custom_data": {
          "bureau": "Experian",
          "report_id": "EXP-2024-10-15-A1B2"
        }
      },
      {
        "credit_report_first_name": "JANE",
        "credit_report_last_name": "DOE",
        "credit_report_request_date": "2024-10-15",
        "credit_score": "712"
      }
    ],
    "property": {
      "property_house_number": "16039",
      "property_street": "PENWOOD DR",
      "property_city": "TAMPA",
      "property_state": "FL",
      "property_zip_code": "33647",
      "property_county": "HILLSBOROUGH",
      "custom_data": {
        "apn": "U-12-28-19-0L8-000019",
        "flood_zone": "X",
        "hoa_monthly_fee": "85.00"
      }
    },
    "compliance": {
      "precise_id_auth_decision": "",
      "precise_id_prove_id_score": "",
      "precise_id_applicant_first_name": "",
      "precise_id_applicant_last_name": "",
      "precise_id_ssn": "",
      "precise_id_score": "",
      "precise_id_decision": "",
      "custom_data": {
        "ofac_check_status": "PASS",
        "cip_review_outcome": "CLEARED"
      }
    },
    "application_notes": [
      {
        "application_note_timestamp": "2024-11-05T14:30:00Z",
        "application_note_text": "---------- OFAC Check (PASS) ---------- ...",
        "application_note_user": "SYSTEM"
      },
      {
        "application_note_timestamp": "2024-11-04T10:21:00Z",
        "application_note_text": "Disclosures sent via DocuSign. BSA Docs & USPS OnBased. VALID IDS:YES",
        "application_note_user": "jsmith"
      },
      {
        "application_note_timestamp": "2024-11-03T09:15:00Z",
        "application_note_text": "This HELOC application is conditionally approved. ...",
        "application_note_user": "jsmith"
      },
      {
        "application_note_timestamp": "2024-11-01T08:45:00Z",
        "application_note_text": "Work flow started",
        "application_note_user": "SYSTEM"
      }
    ]
  }
}
{
  "success": true,
  "loan_id": "LOAN-2025-001234",
  "external_loan_id": "001234-02",
  "friendly_name": "core",
  "source_system": "LOS",
  "extraction_timestamp": "2025-01-15T11:00:00Z",
  "mapped_data": {
    "source": "LOS",
    "source_loan_id": "001234-02",
    "version": "2.0.0",
    "custom_data": {
      "channel": "branch",
      "branch_id": "TPA-04"
    },
    "loan_info": {
      "loan_type": "HELOC",
      "loan_application_id": "001234-02",
      "account_number": "001234",
      "application_id": "02",
      "app_date": "2024-11-05",
      "loan_amount": "42900.00",
      "appraised_amount": "564000.00",
      "ltv_ratio": "31.638%",
      "new_expense_ratio": "290.970%",
      "occupancy_type": "Primary Residence",
      "purpose_of_loan": "Home Improvement",
      "disclosures_date": "2024-11-04",
      "due_date_for_disclosures": "2024-11-08",
      "loan_officer_name": "John Smith",
      "loan_officer_email": "jsmith@examplecu.com",
      "assigned_user": "Mary Johnson",
      "assigned_user_email": "mjohnson@examplecu.com",
      "qualifying_credit_score": "712",
      "custom_data": {
        "internal_promo_code": "REFI-2024-Q4",
        "underwriter_notes_ref": "UW-7741"
      }
    },
    "applicant": [
      {
        "applicant_type": "Main Applicant",
        "first_name": "JOHN",
        "middle_name": "A",
        "last_name": "DOE",
        "applicant_email": "john.doe@example.com",
        "ssn": "589050676",
        "date_of_birth": "1969-03-27",
        "applicant_street": "16039 PENWOOD DR",
        "applicant_city": "TAMPA",
        "applicant_state": "FL",
        "applicant_zip": "33647-1139",
        "occupancy_status": "Own",
        "employer_name": "GINGER HEALTH",
        "position": "PHYSICAL THERAPY",
        "employment_status": "Employed",
        "years_employed": "22",
        "monthly_income": "3511.11",
        "custom_data": {
          "employer_ein": "12-3456789",
          "los_applicant_id": "APP-1001"
        }
      },
      {
        "applicant_type": "Co-Borrower",
        "first_name": "JANE",
        "middle_name": "B",
        "last_name": "DOE",
        "ssn": "412738955",
        "date_of_birth": "1971-08-14",
        "applicant_street": "742 EVERGREEN TERRACE",
        "applicant_city": "ORLANDO",
        "applicant_state": "FL",
        "applicant_zip": "32801",
        "occupancy_status": "Rent",
        "employer_name": "ORLANDO REGIONAL MEDICAL CENTER",
        "position": "REGISTERED NURSE",
        "employment_status": "Employed",
        "years_employed": "12",
        "monthly_income": "5200.00"
      }
    ],
    "credit_report": [
      {
        "credit_report_first_name": "JOHN",
        "credit_report_last_name": "DOE",
        "credit_report_request_date": "2024-10-15",
        "credit_score": "750",
        "custom_data": {
          "bureau": "Experian",
          "report_id": "EXP-2024-10-15-A1B2"
        }
      },
      {
        "credit_report_first_name": "JANE",
        "credit_report_last_name": "DOE",
        "credit_report_request_date": "2024-10-15",
        "credit_score": "712"
      }
    ],
    "property": {
      "property_house_number": "16039",
      "property_street": "PENWOOD DR",
      "property_city": "TAMPA",
      "property_state": "FL",
      "property_zip_code": "33647",
      "property_county": "HILLSBOROUGH",
      "custom_data": {
        "apn": "U-12-28-19-0L8-000019",
        "flood_zone": "X",
        "hoa_monthly_fee": "85.00"
      }
    },
    "compliance": {
      "precise_id_auth_decision": "",
      "precise_id_prove_id_score": "",
      "precise_id_applicant_first_name": "",
      "precise_id_applicant_last_name": "",
      "precise_id_ssn": "",
      "precise_id_score": "",
      "precise_id_decision": "",
      "custom_data": {
        "ofac_check_status": "PASS",
        "cip_review_outcome": "CLEARED"
      }
    },
    "application_notes": [
      {
        "application_note_timestamp": "2024-11-05T14:30:00Z",
        "application_note_text": "---------- OFAC Check (PASS) ---------- ...",
        "application_note_user": "SYSTEM"
      },
      {
        "application_note_timestamp": "2024-11-04T10:21:00Z",
        "application_note_text": "Disclosures sent via DocuSign. BSA Docs & USPS OnBased. VALID IDS:YES",
        "application_note_user": "jsmith"
      },
      {
        "application_note_timestamp": "2024-11-03T09:15:00Z",
        "application_note_text": "This HELOC application is conditionally approved. ...",
        "application_note_user": "jsmith"
      },
      {
        "application_note_timestamp": "2024-11-01T08:45:00Z",
        "application_note_text": "Work flow started",
        "application_note_user": "SYSTEM"
      }
    ]
  }
}

Output fields

Field	Type	Description
`success`	boolean	Whether parsing completed successfully
`loan_id`	string	Loan identifier from the LOS
`external_loan_id`	string	Source-system loan identifier used to trace the parsed record back to the LOS
`friendly_name`	string	Logical name for the source
`source_system`	string	Source-system name
`extraction_timestamp`	datetime	When the data was extracted from the source
`mapped_data`	object	Data mapped to well-known schema fields
`mapped_data.loan_info.qualifying_credit_score`	string	Three-digit qualifying credit score calculated by the LOS from all applicants and credit providers
`mapped_data.applicant[].applicant_email`	string	Optional applicant email address
`mapped_data.<section>.custom_data`	object	Optional source-specific fields. Supported at `mapped_data`, `loan_info`, each `applicant[]` item, each `credit_report[]` item, `property`, and `compliance`

Enum-constrained fields

Map values from your source system to the following canonical strings. Other values fail contract validation after the Data Parse RPA returns.

Field	Allowed values
`mapped_data.loan_info.loan_type`	`Auto`, `CommercialRealEstate`, `HELOC`, `Mortgage`
`mapped_data.loan_info.occupancy_type`	`Primary Residence`, `Second Home`, `Investment Property`, `Manufactured Home`, `Hobby Farm`
`mapped_data.applicant[].applicant_type`	`Main Applicant`, `Co-Borrower`, `Non-Occupying Co-Borrower`
`mapped_data.applicant[].occupancy_status`	`Own`, `Rent`
`mapped_data.applicant[].employment_status`	`Employed`, `Self-Employed`, `Part-Time`, `Unemployed`, `Retired`, `Other`

Document Retrieval RPA Contract

Retrieves the loan's documents from the document repository, including OnBase, FileNet, and SharePoint, and writes them to the Orchestrator storage bucket. Loan Setup provides the target bucket and the loan's directory prefix; the RPA uploads every document for the loan under that prefix, preserving each file's own name.

Loan Setup then scans the prefix to build the document manifest. There is no separate metadata or index step.

Input fields

Field	Type	Required	Description
`version`	integer	Yes	Contract version for compatibility checking
`loan_id`	string	Yes	Loan identifier for correlation
`target_bucket_name`	string	Yes	Orchestrator storage bucket to upload the files to
`target_prefix`	string	Yes	The loan's directory prefix within the bucket. Every document for this loan belongs under this prefix, and the RPA must not write outside it.
`target_folder`	string	No	Orchestrator folder key where the target bucket lives

RPA input example

{
  "version": 2,
  "loan_id": "LOAN-2025-001234",
  "target_bucket_name": "LoanSetup",
  "target_prefix": "LOAN-2025-001234"
}
{
  "version": 2,
  "loan_id": "LOAN-2025-001234",
  "target_bucket_name": "LoanSetup",
  "target_prefix": "LOAN-2025-001234"
}

Output fields

Field	Type	Description
`version`	integer	Contract version, echoed from the input
`loan_id`	string	Loan identifier, echoed from the input
`status`	enum	Overall outcome: `success` or `failed`
`uploaded`	array	File names uploaded under `target_prefix`

Loan Setup reconciles the uploaded list against the files it finds under target_prefix, which remain the source of truth.

Note:

An empty uploaded list (no files) is a valid, non-error outcome: Loan Setup records that there was nothing to extract or evaluate for the loan and completes the run.

RPA output example

{
  "version": 2,
  "loan_id": "LOAN-2025-001234",
  "status": "success",
  "uploaded": [
    "Application_Form.pdf",
    "Credit_Report.pdf",
    "Appraisal.pdf"
  ]
}
{
  "version": 2,
  "loan_id": "LOAN-2025-001234",
  "status": "success",
  "uploaded": [
    "Application_Form.pdf",
    "Credit_Report.pdf",
    "Appraisal.pdf"
  ]
}

Browser application access and scopes

The Loan Setup web application uses a non-confidential external application because it runs in the user's browser and cannot safely store a client secret. Users sign in with their own UiPath accounts through the OAuth 2.0 Authorization Code flow with Proof Key for Code Exchange (PKCE). The application receives a short-lived user access token that represents the signed-in user; it does not use a shared service account.

The scopes below define the maximum APIs and operations the application can request. A scope is a ceiling, not a permission grant: each signed-in user can access only the resources and actions allowed by their UiPath roles and folder permissions.

Orchestrator scopes:

Scope	Purpose
`OR.Execution`	Trigger and monitor process executions
`OR.Folders`	Access folder structure
`OR.Jobs`	Read and manage job state
`OR.Users.Read`	Read user and group details needed to identify and display the signed-in user

Data Fabric scopes:

Scope	Purpose
`DataFabric.Data.Read`	Read agent outputs and loan data from Data Fabric
`DataFabric.Data.Write`	Write agent findings to Data Fabric
`DataFabric.Schema.Read`	Read Data Fabric schema definitions

Traces API scopes:

Scope	Purpose
`Traces.Api`	Access Maestro workflow execution traces for debugging and audit

Solutions Management scopes:

Scope	Purpose
`Solutions.Deployments.Read`	Read which solution version is deployed; this scope cannot deploy, activate, or uninstall a solution

Infrastructure requirements

The UiPath Robot machines executing the following RPA contracts must be running UiPath Robot version 25.10.x or higher:

DataRetrieval: retrieves raw loan data from source systems such as core banking and LOS
DataParse: transforms raw source data into the normalized schema expected by Maestro agents

This version requirement supports RPA input payloads larger than 10 KB, which are common when retrieving full loan records from core banking systems.

Was this page helpful?

PREVIOUSLoan detail reference

NEXTQuality Control (QA/QC)

Overview​

RPA registry pattern​

Process Trigger Contract​

Authentication and invocation​

Required registration​

Token request​

Trigger invocation​

Mark Loan as Completed Contract​

Data Retrieval RPA Contract​

Data Parse RPA Contract​

Enum-constrained fields​

Document Retrieval RPA Contract​

Browser application access and scopes​

Infrastructure requirements​

Related topics​

Was this page helpful?

Overview

RPA registry pattern

Process Trigger Contract

Authentication and invocation

Required registration

Token request

Trigger invocation

Mark Loan as Completed Contract

Data Retrieval RPA Contract

Data Parse RPA Contract

Enum-constrained fields

Document Retrieval RPA Contract

Browser application access and scopes

Infrastructure requirements

Related topics