Document Analyser and Fraud Detection Overview

Overview

The Effi Document Analyser automatically validates and reviews client-uploaded documents for accuracy, completeness, compliance, and potential fraud.

It combines document verification with AI-powered fraud detection to ensure files are authentic, correctly formatted, and properly stored.

Once triggered, the system checks for:

• Expiry dates
• Required time periods
• Correct document types
• Applicant matching
• Naming convention compliance
• Potential document tampering or manipulation
• Financial and ABN anomalies

Validated documents are automatically renamed and filed into SharePoint according to your broker group’s structure.

This improves document quality, reduces follow-ups, enhances compliance, and minimises fraud risk.

Objectives

This guide outlines how to:

• Analyse submitted documents using AI
• Detect potentially fraudulent files
• Review validation and fraud results
• Take appropriate action when issues are found

Supported Document Types

Common supported documents include:

• Payslips
• Bank statements
• Tax and financial documents
• Employment letters

How It Works

1. Upload Required Document

Upload a payslip or other supported document types

Documents can be uploaded via:

• Client (Consumer) Portal
• Broker Portal

• Upload the required documents (e.g., passport, driver's license, etc). 

Step 2: Access Submitted Documents

1. Open the Lead or Deal.
2. Navigate to Documents.
3. Click View Request to see requested files.
4. Review all submitted documents.

Step 3: Run Document Analysis

Brokers can trigger analysis in two ways:

• Automatically, when the document pack is 100% complete, or

• Manually at any time

   

To run manually:

• Open the deal.
• Go to the Documents tab.
• Click 'Analyse with AI'.
• Wait for the process to complete.

The analyser will:

• Validate document rules
• Apply fraud detection checks
• Flag issues
• Auto-rename valid files
• File approved documents into SharePoint

What Happens After Analysis

After processing is complete:

• Passed documents are renamed and filed in SharePoint
• Warnings remain available for broker review
• Rejected documents require client replacement

All results remain visible in the Documents tab for auditing and tracking.

Key Features

4.1 Automated Document Checks

The system validates:

• Expiry (e.g., IDs, passports)
• Document type matching
• Required date ranges (e.g., payslips, bank statements)
• Incorrect or mismatched uploads
• Applicant matching (Primary vs Secondary)
• Naming convention compliance

4.2 New: Fraud Detection Layer

Fraud detection adds advanced analysis to identify suspicious or manipulated documents.

It evaluates files using multiple AI layers and assigns a fraud score.

Common indicators include:

Edited Payslips

• Adjusted salary figures
• Modified year-to-date contributions
• Inconsistent totals

Document Tampering

• Adobe Acrobat or Word overlays
• Font changes
• Formatting inconsistencies
• Mismatched calculations

ABN Verification Failures

• Invalid ABNs
• Unverified business details

Multiple Anomalies

• Documents with several issues receive higher fraud scores

4.3 Auto-Rename

• Applies your group’s naming conventions
• Ensures consistent and compliant file naming
• Reduces manual renaming

4.4 Auto-File to SharePoint

• Moves validated documents to the correct folders
• Uses configured folder structures
• Reduces manual administration

4.5 Manual Trigger

• Brokers can re-run the analysis at any time
• Useful when clients upload new documents mid-process

Rule Configuration

Rules can be customised per document category.

Fraud Scoring & Indicators

Fraud Shield Icon

• A shield icon appears on the analysed documents
• Indicates fraud detection has been applied

Fraud Score

• Displayed as a percentage (e.g., 90%, 93%)
• Higher scores indicate greater risk

Viewing Details

Hover over the shield icon to view:

• Fraud score
• Classification reasons
• Detected anomalies

Examples include:

• Under-contributed superannuation
• Inconsistent income reporting
• Invalid ABN details

Result Types

Reviewing Rejected Documents

When a document is rejected:

1. Open the document record.
2. Review the rejection notes.
3. Check validation and fraud details.
4. Confirm whether the issue is legitimate.
5. Request a replacement if required.

Rejection notes provide clear explanations for transparency.

Override Automated Decisions (if necessary)

If you believe an automated decision is incorrect:

1. Review all analysis details.
2. Confirm supporting evidence.
3. Apply an override if justified.
4. Document the reason for audit purposes.

⚠️ Use overrides cautiously and only with valid justification.

Responsibilities

• Brokers: Review warnings and rejections, request replacements
• Support/Admin: Configure rules, SharePoint, and system settings
• Clients: Re-upload rejected documents

Setup & Configuration

Ensure the following are active:

• SharePoint integration
• Folder structure configuration
• Naming conventions
• Auto-rename feature

Analysis can be set to:

• Run automatically at 100% completion, or

• Run manually at any stage

   

Pricing

• $1 per document analysed
• No per-page charges

• Includes validation and fraud detection

   

Cautionary Notes

• Ensure all required documents are submitted before analysis
• Do not override high-risk fraud flags without evidence

• Maintain audit records for compliance

   

Tips for Efficiency

• Review common rejection reasons
• Analyse multiple leads while waiting for results
• Use warnings as early indicators

• Escalate high fraud scores promptly

   

Need Help or Want to Enable This Feature?

To activate the Document Analyser and Fraud Detection or request assistance, contact:

📩 support@effi.com.au