Scoring Filter Strategy

The OnSefy API provides two key indicators to help assess the trustworthiness of incoming user data:

• risk_level – A simplified label for quick decisioning
• risk_score – A numeric value offering more precision and flexibility

You can use either independently or together depending on your integration needs.

🔹Risk Level#

risk_level - a straightforward categorization:

Use this when you want to act quickly based on level.

🔹Risk Score#

risk_score - a more nuanced scoring range:

• Ranges from 0 to 10
• Lower = safer
• Higher = riskier

You can define thresholds based on your own tolerance levels:

0 - 2      = ✅ Auto-approve
3 - 5     = 🟡 Queue for manual review or soft challenge
5 and above = ❌ Restrict or block immediately

This approach is ideal for building dynamic and adaptive workflows.

🔹 Scoring Intelligence#

The risk_score is powered by an internal scoring engine that considers a wide variety of behavioral and contextual signals.

While the exact scoring mechanics are proprietary, the system evaluates:

• Consistency across submitted values
• Use of uncommon or temporary attributes
• Behavioral anomalies
• Known indicators associated with suspicious activity

Each parameter is assessed in real-time and contributes to the overall trust score using a well-tested scoring methodology.

🔹 Suggested Use#

• Use risk_level if you want to get started quickly with rule-based filters.
• Use risk_score when you need greater flexibility, layered trust logic, or score-based routing.