Real-time fraud detection
Advanced ML algorithms detect and block fraudulent transactions in milliseconds
- Machine learning models
- Pattern recognition
- Behavioral analysis
- Device fingerprinting
Risk Engine
Our Risk Engine minimizes fraud losses by up to 85%
Real-time AI protection that adapts to emerging threats automatically while preserving legitimate customer experiences.
Works with Stripe, Checkout, Adyen, Airwallex, Shopify.
85%
Reduction in fraud losses
<100ms
Average decision time
0.2%
False positive rate
99.9%
System uptime
Security matrix
Comprehensive threat protection
Our Risk Engine protects against all major fraud vectors with industry-leading detection rates.
- Card-not-present fraud
- 99.2%
- Account takeover
- 98.8%
- Synthetic identity fraud
- 97.5%
- Chargeback fraud
- 95.3%
- Friendly fraud
- 89.7%
- Bot attacks
- 99.9%
Engine capabilities
Advanced fraud detection capabilities
Powered by cutting-edge AI and machine learning algorithms that evolve with emerging threats.
Adaptive AI models
Self-learning algorithms that improve fraud detection accuracy over time
- Continuous learning
- Model optimization
- False positive reduction
- Custom rule engines
Instant decision making
Sub-100ms response times for real-time transaction processing
- Ultra-low latency
- High availability
- Scalable infrastructure
- Global edge deployment
Comprehensive monitoring
Complete visibility into your transaction risk landscape
- Risk scoring dashboard
- Threat intelligence
- Compliance reporting
- Audit trails
Deployment path
How it works
From integration to deployment, our streamlined process gets you protected fast.
- 1
Corgi Integration
1-2 daysConnect your payment systems and historical data to Corgi's platform
- 2
Backtested Initial Model
3-5 daysBuild custom fraud models using your historical data with performance estimates
- 3
Live Shadow Testing
15 days15-day comparison testing against existing solutions (Radar, RevenueProtect)
- 4
Corgi Deployment
OngoingFull deployment with continuous monitoring and optimization
Model metrics
Understanding model performance
Learn how precision and recall metrics help optimize your fraud prevention strategy, and how we adjust models to your risk tolerance.
Precision
Non-technical
When we flag a transaction as fraud, how often are we correct? High precision means fewer false alarms and happier customers.
Technical
Precision = True Positives / (True Positives + False Positives). Of all transactions identified as fraud, what percentage actually were fraud.
Example
92% precision = Only 8% false declines
Recall
Non-technical
Of all the actual fraud happening, how much are we catching? High recall means we're not missing fraud attempts.
Technical
Recall = True Positives / (True Positives + False Negatives). Of all actual fraudulent transactions, what percentage did we detect.
Example
89% recall = We catch 89% of all fraud attempts
Adjusting your model strategy
Maximize fraud prevention
Higher Recall: catch more fraud by being more aggressive, accepting some increase in false positives. Best for high-risk industries that prioritize security over convenience.
Reduce false declines
Higher Precision: minimize customer friction by being more selective, though some fraud may slip through. Best for customer-focused businesses prioritizing experience.
Pricing
Choose your analytics package
Flexible pricing that scales with your business growth
Save 16%
MonthlyAnnualCore
For business owners or teams who need advanced analytics to better understand payments and want actionable insights.
$299/month
- Basic payment analytics
- Unlimited analytics (revenue, customer, product, churn, and more)
- Monthly payment insights report
- Email support
- Standard dashboard
- 3 admin seats included, +$15/seat/month for additional seats
- Coming soon: AI analytics agent/chatbot
No credit card required
Most Popular
Pro
For businesses with dedicated payment or risk operations staff who need tools to optimize revenue and prevent fraud.
$999/month
- Everything in Core
- Rule & list management
- Rule intelligence & experimentation
- Weekly payment insights report
- Early fraud warning & intelligent dispute prevention
- 10 user seats with access management, +$10/seat/month for additional seats
- Coming soon: data pipelines & on-demand insights reports
No credit card required
Enterprise
For businesses with multiple PSPs or dedicated payment/operations teams needing enterprise-level security and customized solutions.
Custompricing
- Everything in Pro
- Multi-PSP support
- Enterprise SSO
- Developer API
- Dedicated success manager
- Negotiated seat pricing
No credit card required
Integrates with your existing payments processor via a simple plugin. No development work required. SOC 2 certified.
- AICPA SOC
- Stripe Verified Partner
- Stripe App Marketplace
- Haven Ventures
- Capital X
Get started
Let’s optimize your revenue together
Book a 15-minute demo with our team. We're excited to meet innovative leaders, understand your business model, and help maximize your revenue.