v1.0 · Production · REST API

Detect Fraud.
Before It Happens.

Real-time transaction scoring using an 8-signal weighted ML model. Analyze any financial transaction and get a fraud probability score in milliseconds.

⚡ Get API Key Free Try Live Demo →
8
Risk Signals
<80ms
Avg Latency
25+
Endpoints
99.9%
Uptime
Live Demo

Try it now. Zero setup.

Edit any field and hit Analyze — the real scoring engine runs on your inputs.

User profile: avg $85 · max $250 · 45 txs
Known devices: iPhone 15, MacBook Pro
Known IPs: 192.168.1.1, 10.0.0.5
Known locations: New York, Boston
⚡
Hit Analyze
Results appear here instantly

Running on the live scoring engine — no mock data. Get full API access →

Risk Classification

Four levels of risk

Every transaction is scored 0–1 and classified into one of four risk tiers.

LOW
0.00–0.34
Normal behaviour. Transaction approved automatically.
MEDIUM
0.35–0.59
Mild anomalies. Flag for review or add friction (2FA).
HIGH
0.60–0.79
Strong signals. Webhook alert fired. Manual review recommended.
CRITICAL
0.80–1.00
Transaction blocked. User flagged. Webhook alert fired instantly.
API Reference

Endpoints

All endpoints are under /api/v1. Authenticate via RapidAPI subscription key or Bearer token.

Core Analysis
POST /analyze-transaction Score a transaction — probability, risk level, plain-English explanation All Plans
POST /batch/analyze Score up to 5 (Basic) or 50 (Pro) transactions in one call All Plans
User & Transaction Data
GET /transactions/:user_id Paginated history — last 10 (Basic) or 100 (Pro) per page All Plans
GET /users/:user_id/risk-profile Behavioral profile — avg spend, known devices & locations All Plans
GET /dashboard/stats Platform-wide metrics, 7-day distribution, 30-day trend (Pro) All Plans
Fraud Network Detection
GET /network/device/:fingerprint All users sharing a device — network risk score Pro
GET /network/ip/:address All users sharing an IP — detect shared infrastructure abuse Pro
GET /network/location/:location All users in a location — geography-based ring detection Pro
GET /network/user/:user_id Full network graph for a user across all shared identifiers Pro
Device & IP Reputation
GET /reputation/device/:fingerprint Cross-platform fraud rate for a device fingerprint (anonymised) Pro
GET /reputation/ip/:address Geo + platform fraud-rate lookup using the local GeoLite database Pro
Custom Rules
POST /rules Create a rule — e.g. amount > 5000 AND transaction_type == 'wire_transfer' Pro
POST /rules/test Dry-run a condition against a sample transaction without saving Pro
GET /rules List all active rules with match counts Pro
Velocity Rules
PUT /velocity Set max transactions per hour and max daily spend per user Pro
GET /velocity Retrieve current velocity thresholds Pro
Webhooks
POST /webhooks Register a URL to receive HMAC-signed HIGH/CRITICAL alerts Pro
GET /webhooks List registered webhooks with delivery stats Pro
Outcomes & Calibration
POST /transactions/:id/outcome Submit a chargeback result — auto-adjusts future scoring sensitivity Pro
GET /outcomes/stats Chargeback rate, false-negative rate, current score-boost factor Pro
Export & Analytics
GET /export/transactions/:user_id Download up to 1,000 transactions as a CSV — supports date range filter Pro
GET /merchants/:name/stats Fraud rate, risk distribution, and trend data for any merchant Pro
User Management
POST /users/:user_id/recalibrate Rebuild behavioral baseline from clean transactions only Pro
POST /users/:user_id/unblock Manually clear a blocked user after fraud review Pro
Scoring Model

8 signals. Weighted. Real-time.

Each transaction is evaluated across 8 independent risk signals, combined into a single probability score.

💰
Amount Anomaly 25%
vs. user's historical average spend
📱
New Device 15%
device not seen in user's history
📍
Location Anomaly 15%
location outside known patterns
⚡
High Frequency 15%
burst of transactions in 10 minutes
🌐
New IP Address 10%
IP not in user's known list
🌙
Night Activity 10%
transaction between 1–5 AM UTC
🔺
High Amount 5%
absolute thresholds $5k / $10k / $20k+
🔀
Transaction Type 5%
wire transfers & P2P score highest
Pro Plan

Built for teams that take fraud seriously.

The core engine scores every transaction. Pro adds the intelligence layer on top.

Scoring
Know exactly why it flagged
Instead of a number to decode, you get a sentence with the actual values — "$5,000 is 33x this user's average spend, from a device and location never seen before." Share it with your ops team as-is.
Detection
Catch rings, not just individuals
When the same device or IP turns up across multiple accounts, we surface the connection. Find coordinated fraud before it spreads — one query, full network map.
Intelligence
Know who's really behind the request
Resolve location context locally and combine it with cross-platform fraud rates for the same IP, without depending on a request-time third-party lookup service.
Rules
Your business logic, enforced automatically
Write a one-liner like amount > 5000 AND transaction_type == 'wire_transfer' and we block or escalate it on every transaction going forward. Test it against live data before saving.
Limits
Stop burst attacks before they land
Cap any user at a transaction count per hour or a spend ceiling per day. The moment they breach it, the transaction is escalated to CRITICAL — no extra code on your end.
Alerts
Get notified before your customer does
Register a URL and we'll POST a signed payload the moment a HIGH or CRITICAL transaction hits. HMAC-SHA256 signed so your server can verify it's genuinely from us.
Feedback
The model gets smarter every time you report a chargeback
Tell us when a LOW-scored transaction turned into a dispute. We quietly raise your scoring sensitivity so the same pattern doesn't slip through twice.
Analytics
See which merchants are a problem
Pull fraud rates, risk breakdowns, and the most recent incidents for any merchant — across every user on the platform, without exposing anyone's identity.
Export
Your data in a spreadsheet, on demand
One request gives you up to 1,000 transactions as a clean CSV — with scores, outcomes, and chargeback flags included. Filter by date range. Finance teams love this.
Pricing

Simple, transparent pricing.

Start free. Upgrade when you need the full feature set.

Basic
Free
10 req/min · 500 req/month
  • ✓ Fraud probability score
  • ✓ Risk level (LOW / MEDIUM / HIGH / CRITICAL)
  • ✓ Reason breakdown
  • ✓ Batch scoring (up to 5/call)
  • ✓ Transaction history (10/page)
  • ✓ User risk profile
  • ✓ Dashboard stats
  • ○ Score explanations in plain language
  • ○ Per-signal score breakdown
  • ○ Fraud network detection
  • ○ Custom rules & velocity limits
Pro
$49 /month
500 requests / minute
  • ✓ Everything in Basic
  • ✓ Score explanations in plain language
  • ✓ Per-signal score breakdown
  • ✓ Batch scoring (up to 50/call)
  • ✓ Transaction history (100/page)
  • ✓ Fraud network detection (device/IP/location)
  • ✓ Device & IP reputation lookup
  • ✓ Custom rule engine
  • ✓ Velocity rules (hourly + daily limits)
  • ✓ Subscriber webhooks (HMAC-signed)
  • ✓ Chargeback outcome tracking & calibration
  • ✓ Merchant risk statistics
  • ✓ CSV export (up to 1,000 rows)
  • ✓ User recalibration & manual unblock
  • ✓ 30-day risk trend analytics

Need higher volume or a custom plan? Contact us and we'll sort it out.