SolidWall WAF's internals
Unsupervised machine learning
We use machine learning to fill the semantic gap between HTTP transport and business logic of your application.
Our approach is to teach WAF to map from raw HTTP data to business logic level protocol and it's data objects, such as: actions and scenarios of actions (usecases). This mapping allows us to validate both the syntax and the dynamic behavior, perform type enforcement for actions parameters and control action sequences.

Here we have an example of raw POST request and how it maps to business logic level entities.
Made on
Tilda