Risk & Fraud Tools

Risk and fraud adapters integrate third-party scoring and decisioning engines into the payment flow. Unlike payment adapters, these do not process money — they provide risk signals that the routing engine uses to accept, challenge, or decline transactions.

4 integrations in this category.

---

Available Integrations

Provider	Adapter name	Type	Coverage	Required params
Sift	sift	ML-based fraud scoring	Global	api_key, account_id
Riskified	riskified	Chargeback guarantee	Global (e-commerce)	shop_domain, auth_token
Signifyd	signifyd	Chargeback guarantee	US, UK, EU	api_key
Forta	forta	On-chain threat intelligence	Blockchain txns	api_key

---

How Risk Tools Work

Risk tools are called before routing in the FlowDirect pipeline. The flow is:

Transaction created
       ↓
Risk check (Sift / Riskified / Signifyd)
       ↓
Score returned → Routing decision
       ↓
  Accept → route to payment processor
  Challenge → trigger additional auth (3DS, OTP)
  Decline → reject with error_description: "risk_declined"

Risk tool results are stored in the transaction's metadata field and available in the Admin Console.

---

Provider Details

Sift

ML fraud prevention platform covering payments, account takeover, content abuse, and disputes.

Use cases: Card-not-present fraud scoring, velocity rules, device fingerprinting.

Signal types returned:

• score (0–100, higher = riskier)
• label (bad / good)
• reasons (array of signals)

Config example:

{
  "adapter": "sift",
  "params": {
    "api_key": "YOUR_API_KEY",
    "account_id": "YOUR_ACCOUNT_ID",
    "score_threshold": 80
  }
}

---

Riskified

Chargeback guarantee product for e-commerce. Riskified takes full chargeback liability for approved orders.

Use cases: E-commerce checkout, high-value orders, travel & ticketing.

Model: Submit order → Approve/Decline decision → If approved and later charged back, Riskified pays.

Config example:

{
  "adapter": "riskified",
  "params": {
    "shop_domain": "yourshop.com",
    "auth_token": "YOUR_AUTH_TOKEN"
  }
}

---

Signifyd

Chargeback guarantee with a focus on North America and Europe. Covers fraud and authorization decline recovery.

Use cases: Retail, digital goods, subscription businesses.

Config example:

{
  "adapter": "signifyd",
  "params": {
    "api_key": "YOUR_API_KEY"
  }
}

---

Forta

On-chain threat intelligence for blockchain transactions. Detects malicious wallet addresses, rug pulls, and on-chain attack patterns.

Use cases: Crypto payment flows, DeFi integrations, blockchain compliance.

Config example:

{
  "adapter": "forta",
  "params": {
    "api_key": "YOUR_API_KEY"
  }
}

---

Configuring Risk Tools on a Terminal

Risk tools are assigned at the terminal level. A terminal can have one primary risk tool and optionally a fallback.

Via Admin Console

1. Go to Terminals → select terminal → Risk Settings
2. Select provider from the dropdown
3. Enter credentials
4. Set score_threshold (Sift) or leave at default for guarantee models
5. Save

Via API

curl -X PUT https://4pay.online/api/v1/admin/terminals/TERMINAL_ID \
  -H "Authorization: Bearer ADMIN_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "params": {
      "risk_adapter": "sift",
      "risk_config": {
        "api_key": "YOUR_SIFT_KEY",
        "account_id": "YOUR_ACCOUNT_ID",
        "score_threshold": 75
      }
    }
  }'

---

Built-in Anti-Fraud (Gen2 AF Engine)

In addition to third-party risk tools, the platform includes a built-in Gen2 Anti-Fraud engine that operates independently of external providers:

Feature	Description
Velocity rules	Limits per card, IP, device, partner, terminal
Amount rules	Per-transaction and daily/monthly caps
Geo rules	Country/region block lists
Card BIN rules	Block by issuer country or card range
Custom rules	Configurable via Admin Console

The Gen2 AF engine runs on all transactions regardless of whether a third-party risk tool is configured.

---

Custom risk integrations

Need to integrate a risk provider not listed here? The adapter framework supports custom risk tool adapters. Contact your implementation team.