Credit Score /Fraud Score API Call

1️⃣ Common Credit Bureaus & Their APIs

Credit Bureau	API / Integration	Notes
CIBIL (India)	CIBIL Score API	Requires partnership; returns score (300–900) and detailed report.
Experian	Experian Credit Score API	Returns credit score, credit history, risk grade.
Equifax	Equifax Credit Score API	Provides score and credit report; supports REST/SOAP.
CRIF High Mark	CRIF API	Mainly for commercial and consumer credit data.
TransUnion	TransUnion API	Provides score and risk assessment.

2️⃣ Typical API Call Flow

Request:

POST /getCreditScore
{
  "customerId": "12345",
  "name": "John Doe",
  "dob": "1990-01-01",
  "pan": "ABCDE1234F",
  "address": "Mumbai, India"
}

Response:

{
  "creditScore": 750,
  "scoreDate": "2025-09-21",
  "reportId": "RPT-987654",
  "riskGrade": "Low"
}

Notes:

• Most APIs require unique identifiers like PAN, SSN, or customer ID.

• Some APIs can also return credit history, delinquencies, and credit utilization.

• API usually requires authentication (API key / OAuth / client certificate).

3️⃣ Integration in Loan Flow

1. Collect customer info from onboarding form.

2. Call credit bureau API with required fields.

3. Receive credit score and report.

4. Pass this score to your Loan Decision Engine along with other inputs:

   • Internal risk/fraud score

   • External fraud checks

   • Income-to-debt ratio, LTV, EMIAffordability

5. Decision:

   • Low risk: Approve

   • High risk: Reject

   • Medium risk: Manual review

For fraud scoring in a banking or lending flow, it’s usually handled by either an external fraud detection service or your internal ML-based engine. The API you call depends on which solution you integrate. Here’s a detailed breakdown:

1️⃣ External Fraud Detection APIs

Provider	API / Integration	Notes
Refinitiv World-Check / Dow Jones	Watchlist Screening API	Checks customer against global sanctions, PEPs, adverse media. Returns match/no-match and risk score.
Actimize (NICE)	Fraud Detection API	Batch or real-time; can score transactions and customer onboarding.
FICO Falcon / FICO Platform	Fraud Score API	Returns risk score for transactions or onboarding.
Experian Fraud Detection	Experian Fraud Score API	Returns a risk score for identity verification.
Internal ML Model	Custom REST API	Accepts customer details, transaction history, behavior data; returns internal fraud probability score.

2️⃣ Typical API Call Flow

Request Example (Internal ML or External API):

POST /getFraudScore
{
  "customerId": "12345",
  "name": "John Doe",
  "dob": "1990-01-01",
  "pan": "ABCDE1234F",
  "accountNumber": "1234567890",
  "transactionHistory": [
    {"txnId": "T1", "amount": 5000, "date": "2025-09-20"}
  ]
}

Response Example:

{
  "internalFraudScore": 0.25,
  "externalFraudScore": 0.70,
  "riskLevel": "Medium",
  "reviewRequired": true
}

Notes:

• internalFraudScore → your ML model prediction (0–1 probability).

• externalFraudScore → from external vendor (0–1 or Low/Medium/High).

• Can trigger manual review if risk is medium/high.

3️⃣ Integration in Loan / Onboarding Flow

1. Collect customer info → call fraud score API.

2. Get internal & external fraud scores.

3. Combine with other inputs like credit score, income, LTV.

4. Pass to Loan Decision Engine.

5. Decision outcome:

   • Low risk: Approve

   • Medium risk: Manual review

   • High risk: Reject

💡 Tip: Fraud scoring often combines multiple sources: internal ML model + external API (watchlist, sanctions, fraud databases). This gives a robust risk assessment.

Experian Hunter is a fraud detection and identity verification solution provided by Experian. In banking and lending, it’s primarily used to detect and prevent fraud during customer onboarding and transactions. Here’s a detailed breakdown:

1️⃣ Primary Purpose

• Identity verification: Checks if the person applying for a loan or account is who they claim to be.

• Fraud detection: Detects suspicious activity, synthetic identities, or identity theft.

• Risk scoring: Assigns a fraud risk score based on multiple signals.

2️⃣ Key Use Cases

Use Case	Description
Digital Onboarding	Validates PAN, Aadhaar, or other government IDs against official databases and fraud databases.
AML / KYC Checks	Flags PEPs (Politically Exposed Persons), sanctions, and adverse media alerts.
Transaction Screening	Monitors high-risk transactions for potential fraud.
Synthetic Identity Detection	Detects fake identities created using stolen or fabricated data.
Behavior Analysis	Checks if user behavior during onboarding is consistent with legitimate customers.

3️⃣ How It Works

1. Collect customer info: Name, DOB, PAN, email, phone, address, device info.

2. Send API request to Experian Hunter:

POST /hunter/checkIdentity
{
  "customerId": "12345",
  "name": "John Doe",
  "dob": "1990-01-01",
  "pan": "ABCDE1234F",
  "email": "john.doe@example.com",
  "phone": "9999999999"
}

3. Receive response with:

{
  "fraudScore": 0.75,
  "riskLevel": "High",
  "alerts": ["PEP match", "Email suspicious"]
}

4. Decision:

   • Low risk → Approve

   • Medium risk → Manual review

   • High risk → Reject

4️⃣ Integration in Loan / Onboarding Flow

• Step 1: Customer fills onboarding form

• Step 2: Call Experian Hunter API for identity & fraud check

• Step 3: Receive fraud score and alerts

• Step 4: Pass scores to Loan Decision Engine along with credit score, internal ML fraud score, LTV, EMIAffordability

• Step 5: Trigger approval, manual review, or rejection

💡 Key Point: Experian Hunter is not a credit score service—it’s purely for fraud detection and risk scoring to prevent onboarding or transaction fraud.

1️⃣ Dow Jones Risk & Compliance / World-Check API

• Purpose: Global sanctions, PEPs (Politically Exposed Persons), adverse media, anti-money laundering (AML) checks.

• Use Case:

  • Onboarding high-risk customers

  • Screening transactions for sanctions / regulatory compliance

  • Monitoring ongoing customer risk

• Integration:

  • REST/SOAP API or SFTP batch

  • Returns match/no-match, risk level, PEP / sanctions alerts

• Response Example:

{
  "customerId": "12345",
  "match": true,
  "riskLevel": "High",
  "alerts": ["Sanction match: XYZ", "PEP match"]
}

2️⃣ Refinitiv / World-Check API

• Purpose: Same as Dow Jones; sometimes more comprehensive data coverage depending on geography and corporate intelligence.

• Use Case:

  • AML compliance checks

  • Fraud / identity risk assessment

  • Regulatory reporting (FATCA, OFAC, UN sanctions, etc.)

• Integration:

  • REST/SOAP API or batch SFTP feed

  • Returns risk scores, match flags, and detailed intelligence notes

• Response Example:

{
  "customerId": "12345",
  "match": false,
  "riskScore": 0.15,
  "alerts": []
}

3️⃣ Key Differences

Feature	Dow Jones	Refinitiv
Coverage	Strong in sanctions, PEP, negative news	Strong in corporate/financial intelligence, global PEP coverage
Data Updates	Daily	Near real-time or daily
Integration	API or batch	API or batch
Licensing	Usually per API call or subscription	Usually per API call or subscription
Global Reach	Strong in US/Europe	Strong in EU/Asia-Pacific coverage

4️⃣ Integration in Lending / Onboarding Flow

1. Collect customer info (PAN, SSN, name, DOB, address, phone).

2. Call Dow Jones or Refinitiv API for sanctions/PEP/adverse media check.

3. Receive risk level / alerts.

4. Combine with:

   • Internal fraud score (ML)

   • External fraud API (Experian Hunter)

   • Credit score (CIBIL / Experian / Equifax)

5. Pass to Loan Decision Engine → Approve / Manual Review / Reject

💡 Key Note: Many banks use both external fraud checks (Experian Hunter / internal ML) and sanctions screening (Dow Jones or Refinitiv) together for a robust compliance & fraud workflow.

1️⃣ Credit Score API

Vendor	Recommendation	Reason
CIBIL	✅ Primary choice in India	Most widely accepted by Indian banks and NBFCs; gives official credit score (300–900) and detailed credit report; used for regulatory and underwriting purposes.
Experian (India)	Optional / secondary	Can be used if customer opts for Experian report or if you want alternative credit bureau data. Useful for cross-checking or in multi-bureau scoring.

💡 Best Practice:

• Use CIBIL as the primary API for loan decision engine.

• Optionally, pull Experian score if you want multi-bureau validation to reduce risk.

2️⃣ Fraud Score API

Vendor	Recommendation	Reason
Experian Hunter	✅ Primary for identity verification & fraud scoring	Best for digital onboarding fraud detection, synthetic ID detection, internal + external risk scoring.
Refinitiv / Dow Jones	✅ For AML / sanction checks	These are regulatory compliance and sanctions checks, not pure fraud detection. Use them in addition to Experian Hunter, not instead.

💡 Best Practice:

• Use Experian Hunter API to calculate internal + external fraud score during onboarding.

• Call Refinitiv or Dow Jones API to check PEPs, sanctions, and adverse media.

• Combine credit score + fraud score + AML/sanctions check in your loan decision engine to assign risk levels (Low / Medium / High).

Summary:

Use Case	API to Call	Notes
Credit Score	CIBIL (primary)	Optional Experian for multi-bureau validation
Fraud / Identity Risk	Experian Hunter	Detects synthetic identity & onboarding fraud
Sanctions / Regulatory Risk	Refinitiv or Dow Jones	AML, PEP, OFAC, FATCA checks

1️⃣ Refinitiv / World-Check

• Purpose: Sanctions, PEP, adverse media, identity risk screening.

• Use Case: Can act as a fallback to detect high-risk individuals, though not a full fraud ML engine like Hunter.

• Pros: Widely recognized for regulatory compliance.

• Cons: Less focused on synthetic identity detection and device/behavioral fraud scoring.

2️⃣ Dow Jones Risk & Compliance

• Purpose: Global sanctions, PEPs, negative media.

• Use Case: Can serve as a fallback for AML/fraud screening during onboarding.

• Pros: Strong global coverage and regulatory acceptance.

• Cons: Mainly checks watchlists; less effective for real-time identity fraud or synthetic identity detection.

3️⃣ Internal ML-Based Fraud Scoring

• Purpose: Custom-built ML model for internal fraud scoring.

• Use Case: If Experian Hunter API fails or is unavailable, your internal model can compute a fraud probability score based on:

  • Device fingerprinting

  • Behavioral analytics

  • Transaction patterns

  • Historical fraud data

• Pros: Fully under your control, real-time scoring.

• Cons: Needs training, ongoing maintenance, and sufficient historical data.

4️⃣ Other Commercial Alternatives

Vendor	Use Case
FICO Falcon / FICO Platform	Transaction and onboarding fraud detection, scoring
SAS Fraud Management	AML + fraud detection for digital banking
Experian India Fraud Solutions (other modules)	Some clients use Experian’s other APIs if Hunter is unavailable

💡 Recommended Fallback Strategy

1. Primary fraud API: Experian Hunter

2. Fallback 1: Internal ML fraud scoring engine

3. Fallback 2: Refinitiv / Dow Jones watchlist check (AML/PEP/adverse media)

4. Combine credit score + fraud score + sanctions check in the Loan Decision Engine to assign Low / Medium / High risk.