Bank Statements — Docs Intelligence

Docs Intelligence — Bank Statement Analysis

Upload bank statements to get a comprehensive Financial Health Profile and Fraud Risk Profile. The system produces scored, reasoned, and explained assessments calibrated specifically for African financial markets.

What You Get

Financial Health Profile

• 4 Dimension Scores (0-100): Revenue, Cash Flow, Expense and Liability, Balance Behaviour — each rated Strong / Adequate / Weak / Insufficient
• 6 African Market Signals: Tontine Intelligence, Mobile Money Shadow Balance, Bancarization Quality, Calendar-Adjusted Stress, Social Capital, Diaspora Income
• Credit Readiness Score (0-100) with credit limit band
• LLM Narrative (3-5 paragraphs) explaining the financial picture
• Recommendations: Additional documents, conditions, monitoring

Fraud Risk Profile

• Fraud Score (0-100) with verdict: Proceed / Proceed with conditions / Refer / Decline
• Individual flags with dimension, pattern, confidence, cross-dimensional corroboration
• Named anomaly clusters: Pre-Application Staging, Circular Flow, Ghost Business, Debt-Driven Distress, Dangerous Payer Concentration
• LLM Fraud Narrative explaining each significant signal

API Endpoints

Upload and Analyze

POST https://api.tryenvoyx.com/api/v1/bank-statements/upload

Request Fields

Streaming

Add X-Stream-Response: true header for real-time Server-Sent Events through all 14 processing steps.

Code Examples

curl -X POST https://api.tryenvoyx.com/api/v1/bank-statements/upload \
  -H "X-API-Key: YOUR_API_KEY" \
  -F "file=@bank_statement.pdf" \
  -F "business_type=healthcare_provider" \
  -F "sector=healthcare"

const formData = new FormData()
formData.append('file', fs.createReadStream('bank_statement.pdf'))
formData.append('business_type', 'healthcare_provider')

const response = await fetch('https://api.tryenvoyx.com/api/v1/bank-statements/upload', {
  method: 'POST',
  headers: { 'X-API-Key': 'YOUR_API_KEY' },
  body: formData,
})

const { data } = await response.json()
console.log('Credit Readiness:', data.bank_statement.fh_credit_readiness_score)
console.log('Fraud Verdict:', data.bank_statement.fraud_risk_verdict)

import requests

with open('bank_statement.pdf', 'rb') as f:
    response = requests.post(
        'https://api.tryenvoyx.com/api/v1/bank-statements/upload',
        headers={'X-API-Key': 'YOUR_API_KEY'},
        files={'file': f},
        data={'business_type': 'healthcare_provider'},
    )

stmt = response.json()['data']['bank_statement']
print(f"Credit Readiness: {stmt['fh_credit_readiness_score']}/100")
print(f"Fraud Verdict: {stmt['fraud_risk_verdict']}")

The Four Dimensions (0-100 each)

Each dimension is scored 0-100 and rated:

Revenue Analysis

Classifies every inflow by source: institutional payments, mobile money sweeps, cash deposits, tontine receipts, related-party transfers, diaspora transfers. Evaluates consistency against what this business type actually earns and on what cycle.

Cash Flow Health

Measures how the business manages gaps between major inflows. In markets where institutional payers clear every 60-90 days, the inter-cycle floor is the real stress test. Identifies mobile money and tontine liquidity bridges.

Expense and Liability Patterns

Identifies what a real operating business should be spending on and whether those expenses are present. Separates tontine from debt service, personal from operational. Assesses total visible debt load including mobile credit apps.

Balance Behaviour

Analyses balance patterns for timing manipulation (pre-application conditioning), survival behaviour between inflows, and trend over the full period.

African Market Signals

Six proprietary signals calibrated for how money actually moves in West Africa:

Credit Readiness Score

The composite score (0-100) is a weighted combination:

Adjusted by: African market signal bonus/penalty (up to 5 points), legitimacy gate (cap at 20 if document not legitimate).

The score maps to a credit limit band calibrated to the currency (e.g. "500K-2M XOF").

Fraud Risk Score and Verdict

Named Anomaly Clusters

When flags from multiple dimensions point to the same pattern, they are grouped into named clusters:

• Pre-Application Staging — fabricated activity near application date
• Circular Flow / Manufactured Inflow — coordinated deposit rings
• Ghost Business — missing expected operating expenses
• Debt-Driven Distress — loan stacking across multiple lenders
• Dangerous Payer Concentration — over-reliance on single income source

Processing Pipeline (14 Steps)

PDF Report

Every analyzed statement generates a PDF report accessible via:

GET https://api.tryenvoyx.com/api/v1/bank-statements/{id}/report

Returns a presigned URL for the PDF. The report includes both profiles, all scores, narratives, flags, anomaly clusters, and recommendations — formatted for credit committee review.

Webhook Events

See Webhooks for setup and WebSockets for real-time streaming.

Preferences API

Set defaults so you do not pass business_type and sector with every upload.

curl -X PUT https://api.tryenvoyx.com/api/v1/bank-statements/preferences \
  -H "X-API-Key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"default_business_type": "healthcare_provider", "default_sector": "healthcare"}'

await fetch('https://api.tryenvoyx.com/api/v1/bank-statements/preferences', {
  method: 'PUT',
  headers: { 'X-API-Key': 'YOUR_API_KEY', 'Content-Type': 'application/json' },
  body: JSON.stringify({ default_business_type: 'healthcare_provider', default_sector: 'healthcare' }),
})