Lead Scoring Validation with Fraud Prevention
Sales teams need reliable lead scores, but AI models can be gamed by fake leads or adversarial input that inflates scores.
El problema
Lead scores produced by AI models are validated by Corules before routing to sales. Email domain reputation, phone number validity, IP geolocation consistency, and submission velocity checks filter out fraudulent leads. Suspicious leads are flagged for manual review instead of auto-routing to sales, preserving pipeline quality.
Cómo lo resuelve Corules
ESCALATE: IP geolocation (Nigeria) does not match stated country (United States). Flagged for manual review.
Ejemplo de política
// Lead validation policy (CEL)
context.email_domain_reputation >= params.min_domain_reputation
&& context.phone_valid == true
&& context.ip_geo_matches_country == true
&& context.submission_velocity_score < params.max_velocity_scoreOpciones de integración
SalesforceREST API
Frequently Asked Questions
Does this replace AI lead scoring?
No. Corules validates the output of the scoring model, not the score itself. The AI scores; Corules confirms the input data is legitimate before the score is acted upon.