Technical Integration Guide: Flinks Upload (Document Processing)

Flinks Upload is an essential component of the Flinks platform, designed to process, authenticate, and extract critical financial data from customer-provided documents, such as bank statements and void cheques. It offers an effective alternative or fallback mechanism to traditional direct bank linking, reducing drop-off rates and streamlining underwriting processes by automating document review and robust fraud detection.

This guide focuses on the technical integration of Flinks Upload, covering implementation via Flinks Connect (the client-facing interface) and data access/review via the Client Dashboard and supporting Flinks API endpoints.

I. Flinks Upload Implementation via Flinks Connect (Front-End)

Flinks Upload is designed to seamlessly integrate into the customer journey via the Flinks Connect iframe, often used to connect customer bank accounts. Enabling this feature requires contacting a Flinks Representative.

A. Integration Modes

Flinks Connect supports flexible document upload configurations based on desired workflow primacy:

1. Primary Option: The user is immediately presented with the choice to either link their bank account or upload statements immediately after selecting their financial institution. This mode inherently includes the fallback logic (e.g., if a direct connection fails).

2. Fallback Option: Document upload is presented only if the direct bank connection fails due to issues such as a failed login, wrong credentials, account inactivity, or if the institution is unlisted.

Alternatively, customers can be sent directly to an upload-only flow by appending /Upload to the Flinks Connect URL.

B. Flinks Connect Parameters and Configuration

Integration is managed through URL parameters embedded in the Flinks Connect iframe URL.

Note on File Types: Flinks Upload supports PDF documents and scanned image formats, including JPEG, PNG, and TIFF. The upload limit is a maximum of 15 bank statements, or up to 12 months of financial documents per account/financial institution, with a maximum size of 180 MB total or 20 MB per file.

C. Handling Upload Events via JavaScript Listener

For greater control over the user experience and tracing purposes, developers should enable the JavaScript Event Listener. Flinks Connect triggers specific events related to the document upload process:

If an event like UPLOAD_INVALID_FILE is triggered, further backend action or immediate user messaging may be required.

II. Integration via Client Dashboard and API (Back-End)

Once files are uploaded, processing and data retrieval are handled through the Flinks ecosystem, utilizing the Client Dashboard for operational review and APIs for programmatic access.

A. Uploading Files via the Client Dashboard

The Client Dashboard offers a manual interface for uploading documents by authorized personnel.

Steps to Upload via Dashboard:

1. Navigate to the INSIGHTS screen and select the External Data tab.

2. Click Import PDF (or Import PDF/Image).

3. Select files for a singular customer at a singular financial institution. If uploading multiple customer files or files from multiple institutions, perform separate uploads.

4. Enter a unique identifier (e.g., customer name) for identification when data is delivered.

5. Click Import Files.

Data processing typically finishes within 10 minutes of upload, though it may take up to 90 minutes or four hours in some cases.

B. Receiving and Reviewing Processed Data

Processed data can be accessed through both the Dashboard and API calls.

1. Client Dashboard Review:

To review the extracted data and insights:

1. Navigate to the INSIGHTS screen in the Dashboard.

2. Locate the customer entry and confirm the Request Status is Completed.

3. Click the row to view the processed data. If other services, such as Attributes, are enabled, that data will also be accessible here.

2. Document Fraud Check (Dashboard):

The Dashboard provides a dedicated tool for reviewing document authentication results.

1. From the customer row on the INSIGHTS screen, click on the Document Check tab.

2. This tab lists uploaded files, displaying the fraud verdict and document quality.

3. Clicking on a file allows review of the underlying risk and trust signals identified by Flinks, which checks over 500 fraud signals per document. Fraud verdicts include Trust, Normal, Warning, or High Risk.

C. API Consumption and Fraud Analysis

The data processed from the uploaded files is returned in Flinks' standard data formats. To retrieve this data, follow the standard API flow used for connected accounts (aggregation), utilizing the relevant login identifier.

1. Data Retrieval Flow:

Although the documents are uploaded via the frontend (Flinks Connect) or Dashboard, the data is retrieved by performing API calls that rely on the established loginId.

1. Get LoginId: The system must save the unique, non-expiring loginId issued after a successful upload/connection process.

2. Initiate Session: Call the /Authorize endpoint, specifying the loginId and MostRecentCached:true to retrieve the processed data without initiating a live connection. This exchanges the loginId for a requestId.

3. Retrieve Data: Use the resulting requestId to call aggregation endpoints such as /GetAccountsDetail or /GetAccountsSummary. The API Reference includes dedicated endpoints for handling uploaded data, such as /Upload (Base Example) post and /Upload (All Attributes Example) post.

2. Fraud Detection API Integration:

Fraud Detection Flags derived from statement uploads are available via API. Customers have the option to call the dedicated /FraudAnalysis endpoint to check for fraud signals on the uploaded content. Alternatively, Flinks can be configured to deliver fraud alerts automatically via Webhooks.

• If an integration uses Webhooks for fraud alerts, the API calls for the raw aggregated data can potentially serve as a trigger for subsequent Attribute API calls, although Attributes payloads themselves are generally not delivered via webhooks.