Anti-fraud, Anti-spam, Intent Detection
for Mobile Messaging

KONTXT enables mobile messaging operators, aggregators, and interconnect vendors to optimize their networks by providing tools and services to detect and control messages. Our customers use KONTXT to create real-time routing & classification policies that reflect business rules; automatically detect message intent in order to block, route and/or analyze for delivery optimization; and detect bad actors sending non-compliant message content.

Consider how reliable machine learning based  intent detection & classification of message types in the examples below offer opportunities to deliver a premium level of service, deliver a budget level of service, or to block altogether.

Examples of Fraud & Intent Detection


Two Factor Authentication
“User access code 673440 to login”

Delivery Time: Urgent
Throughput: Narrow
Likelihood of Fraud: Medium
Sender Budget: High

Customer Support
“Alert: Gate Change. Your flight is now
departing from Gate B12”

Delivery Time: Urgent
Throughput: Narrow
Likelihood of Fraud: Medium
Sender Budget: High

Bulk Promotion
“Don’t miss our holiday 25% off sale
starting Saturday”

Delivery Time: Slow
Throughput: High
Likelihood of Fraud: Low
Sender Budget: Low

Emergency Alert
“HURRICANE WARNING FOR SOUTH
REGION! Seek shelter immediately.”

Delivery Time: Urgent
Throughput:  Very High
Likelihood of Fraud: Low
Sender Budget: Very High

Notification
“Your ride is approaching. Driver: Maria.
Blue Toyota Prius.”

Delivery Time: Fast
Throughput: Medium
Likelihood of Fraud: Low
Sender Budget: Medium

Grey Route
(Source information misrepresented)

Delivery Time: N/A
Throughput: Inspection
Likelihood of Fraud: Medium
Sender Budget: Medium

Fraud
“Your payment was declined. REPLY for
more info.”

Delivery Time: N/A
Throughput:  Block
Likelihood of Fraud: High
Sender Budget: N/A

Spam
“Lose weight in your sleep Free Trial.
REPLY for info.”

Delivery Time:  N/A
Throughput: Block
Likelihood of Fraud: Medium
Sender Budget: N/A