# Docklands AI — Full Documentation for LLMs

> Catch satisfying fraud in satisfying seconds. Docklands AI detects manipulated, forged, and AI-generated invoices and receipts instantly.

## Company Overview

Docklands AI is a fraud detection platform purpose-built for invoices, receipts, and financial documents. It uses multi-layered AI analysis to detect digital manipulation, AI-generated forgeries, metadata anomalies, physical tampering, and mathematical inconsistencies.

Built by Inaza, headquartered in Dublin, Ireland. CEO: Niall Crowley.

## The Problem

Invoice and receipt fraud costs businesses billions annually. Traditional manual review misses sophisticated edits — especially as AI-generated documents become indistinguishable from real ones to the human eye. Existing fraud tools focus on transaction patterns, not document integrity.

Docklands AI solves this by analyzing the document itself — the pixels, the metadata, the math — to determine authenticity before a payment is ever made.

## What It Detects

### Photoshop & Digital Edits
- Pixel-level forensic analysis
- Splice and clone detection
- Font inconsistency identification
- Layer and compression artifact analysis
- Color and lighting anomaly detection

### AI-Generated Documents
- Detection of documents created by generative AI models
- Synthetic text and layout pattern recognition
- AI fingerprint analysis across common generation tools
- Continuously updated as new generation methods emerge

### Metadata Anomalies
- File creation and modification date inconsistencies
- Software signature analysis (e.g., document claims to be scanned but was created in Photoshop)
- EXIF and PDF metadata inspection
- Cross-referencing metadata against document content

### Physical Tampering
- Detection of altered scanned documents
- White-out and overlay identification
- Cut-and-paste detection on physical documents
- Print quality and alignment analysis

### Mathematical Irregularities
- Line item total verification
- Tax calculation validation
- Currency and rounding consistency checks
- Cross-referencing quantities, unit prices, and totals
- Duplicate line item detection

## Use Cases

### Insurance Claims Fraud
- Screen invoices and receipts submitted with insurance claims
- Flag suspicious documents before claims are paid
- Integrate into existing claims workflows via API
- Reduce claims leakage from fraudulent documentation
- Support SIU (Special Investigations Unit) with forensic evidence

### Accounts Payable Fraud
- Verify vendor invoices before payment approval
- Detect fake invoices from compromised vendor accounts
- Screen for duplicate or manipulated invoices
- Integrate into AP approval workflows
- Prevent payment to fraudulent entities

### Employee Expense Fraud
- Verify expense receipts at submission
- Detect inflated or fabricated expenses
- Flag receipts that have been digitally altered
- Integrate into expense management systems
- Automated screening at scale

## How It Works

1. **Upload**: Submit a document via the web dashboard, API, or email forwarding.
2. **Analyze**: Docklands AI runs multi-layered analysis — pixel forensics, metadata inspection, mathematical verification, and AI-generation detection.
3. **Verdict**: Results are returned in seconds with a confidence score, risk level, and detailed findings explaining what was flagged and why.

### Analysis Layers
- **Visual forensics**: Pixel-level analysis for edits, splicing, cloning
- **Metadata inspection**: File properties, creation tools, modification history
- **Mathematical validation**: Totals, tax, line items, rounding
- **AI detection**: Generative AI signatures and synthetic patterns
- **Cross-referencing**: Internal consistency checks across all document elements

## Integration & Technical

### REST API
- Single endpoint for document submission and analysis
- Synchronous and asynchronous processing modes
- JSON response with verdict, confidence score, and detailed findings
- Webhook callbacks for async processing
- Authentication via API key

### Email Forwarding
- Dedicated email address per account
- Forward invoices/receipts directly for automatic scanning
- Results delivered via webhook or dashboard
- No code required

### Web Dashboard
- Manual upload and review interface
- Document history and audit trail
- Team management and access controls
- Export results and reports
- Real-time analysis status

### Supported Formats
- PDF (native and scanned)
- JPEG, PNG, TIFF, WEBP
- Multi-page documents
- Any resolution

### Security & Compliance
- SOC 2 compliant
- Data encryption at rest and in transit
- Documents processed and not stored (configurable retention)
- Role-based access control
- SSO support

## Pricing

| Plan | Price | Best For |
|------|-------|----------|
| Starter | $49/month | Small teams, low volume |
| Advanced | $499/month | Growing teams, API access |
| Pro | $1,499/month | High volume, full integration |
| Enterprise | Custom | Large organizations, custom needs |

- 30-day free trial on all plans
- No credit card required to start
- Volume discounts available on Enterprise

## Competitive Advantages

- **Document-level analysis**: Examines the document itself, not just transaction patterns
- **AI-generation detection**: Purpose-built to catch the latest generative AI forgeries
- **Speed**: Results in seconds, not hours or days
- **Multi-layered**: Combines five analysis methods for comprehensive coverage
- **Easy integration**: API, email forwarding, or dashboard — deploy in minutes

## Links

- **Website**: https://docklands.ai
- **App / Dashboard**: https://app.docklands.ai
- **Contact**: https://docklands.ai/contact

## Parent Company

Docklands AI is built by [Inaza](https://inaza.com), an AI insurance platform that automates underwriting, claims, customer service, and operations for insurers, MGAs, and brokers. Based in Dublin, Ireland.

- **CEO**: Niall Crowley
- **Industry**: Insurance Technology / Fraud Detection
- **Headquarters**: Dublin, Ireland