Models & Scoring
Understanding Straddle’s risk assessment system, including decision logic, risk scores, and correlation scores
Straddle employs a sophisticated suite of models and scoring systems to evaluate the risk associated with transactions and accounts. This guide provides an in-depth look at our decision logic, risk scores, correlation scores, and KYC field validations.
Decision Logic
The Decision engine is the core component of our risk assessment system, providing real-time recommendations based on a comprehensive analysis of various risk factors.
Key Components
Risk Scores
Evaluates scores from proprietary models
Correlation Scores
Assesses relationships between provided PII elements
KYC Field Validations
Verifies accuracy and completeness of provided PII
Decision Outcomes
When creating an “individual” customer, the possible outcomes are:
pending
verified
review
rejected
resubmit
The Decision Logic can be customized to define specific risk thresholds and decision criteria based on your business requirements.
Accessing Decision Outcomes
You can access the decision outcome when creating or retrieving a customer. Here’s how you can check the status:
Risk Scores
Straddle’s risk scoring models assess the likelihood of fraudulent activity associated with a given set of PII.
Synthetic Identity Risk Score
This score predicts the probability that the provided PII corresponds to a fictitious person.
- Score Range: 0.001 to 0.999
- High Risk Threshold: ≥ 0.900
Fraud Risk Score
This score assesses the risk of identity fraud associated with a transaction.
- Score Range: 0.001 to 0.999
Accessing Risk Scores
You can access the risk scores from the customer response object:
Correlation Scores
Correlation scores measure the strength of the relationship between different elements of the provided PII.
Types of Correlation Scores
- Name-Address Correlation
- Name-Email Correlation
- Name-Phone Correlation
Correlation Categories
Match
Strong Correlation ≥ 0.65
Unknown
Uncertain Correlation 0.20 - 0.65
No Match
No Correlation ≤ 0.20
Accessing Correlation Scores
You can access the correlation scores from the customer response object:
KYC Field Validations
The Decision module performs comprehensive validations on the provided PII fields to ensure their accuracy, completeness, and consistency with trusted data sources.
Validated Fields
- first_name
- last_name
- street_address (displayed as address_line1 in the response)
- city
- state
- zip
- phone
- date_of_birth
- ssn
Validation Results
Field validations return boolean values:
true
: Information correlates with the best-matched entity in trusted data sources.false
: Information does not correlate with the best-matched entity in trusted data sources.
Accessing Field Validations
You can access the field validations from the customer response object:
Conclusion
By leveraging Straddle’s comprehensive risk assessment system, including decision logic, risk scores, correlation scores, and KYC field validations, you can significantly enhance your fraud prevention capabilities and improve the security of your platform. Understanding these components allows for more effective integration and utilization of Straddle’s powerful risk assessment tools in your applications.