Fine-tune detection sensitivity

To evaluate risk, Transmit fraud engine monitors user interactions (e.g. it collects telemetry and user actions) and runs its ML detection models to generate the risk reasons that justify recommendations. The calculation of the risk score relies on the combination of reasons that contribute to the risk assessment.

To meet your specific requirements on risk detection, you can customize the logic behind the risk detection sensitivity by fine-tuning the importance assigned to each type of reasons by the fraud engine. For each type, define its impact on the score calculation choosing from options like ignore, low, default, high, and deny (Admin Portal> detection and response > Configuration > Detection sensitivity).


Before you push your custom detection policy into production, we strongly recommend to save the preview policy configurations and test their effectiveness either by consulting the recommendations' raw data in the Recommendations dashboard or by questioning our Conversational Analytics tool. The fine-tuned configurations won't affect your processes in production until you actively deploy them.

For example, in the event of detecting a sudden surge in bot attacks, an organization can quickly implement ad-hoc detection sensitivity settings, setting the Bot parameter to Deny. This will systematically block all bot activity until a proper strategy is defined.

fine-tune detection sensitivity