The Decision engine view consists of 2 parts:

• Accept/Refuse lists,
• Rules.

Accept/Refuse lists

On the view, you can see all lists configured in your application. Each list should be connected to some "List-based rule". Here you can find the information on what recommendation produce connected rules in case of a match. Depending on the connected rule, we can distinguish three main types of lists:

Refuse list is a list connected to a List-based rule that generates refuse recommendation on a match.

Review list is a list connected to a List-based rule that generates review recommendation on a match.

Accept list is a list connected to a List-based rule that generates either overriding_accept or accept recommendation on a match. This ensures that the transaction will always be accepted and the results of other rules will be ignored.

Creating a new list

By clicking ‘Create a new list’ you can add a new list.

To create a new list it is necessary to input:

• list name,
• the group that it will belong to (Refuse list, Review list or Accept list),
• the list of the fields that will include the elements of the list. Make sure to manually enter desired name of the field and select field type from the dropdown.

You can also create a list when creating a list-based rule, for details, please check List-based rule.

List details

By clicking on a list, you can go directly to the list details and view:

• the list items,
• when the item expires (you can add elements to lists either on a permanent or on a temporary basis by specifying the end date),
• who created the item,
• when the item was created,
• the item's comment.

You can inspect both active and expired elements of lists by navigating between corresponding tabs. You can also delete a single item from the list or use the Batch deleting functionality, which lets you delete multiple points at once. Search bar helps you find specific items.

You can also delete items from lists by clicking ‘Delete button’ on the right. You will be asked for confirmation, before deleting an element.

In the top right corner, you can find the “Add element” button.

You can add multiple items at once. The elements could be also provided to Panel by pasting it from the clipboard, separated by a new line.

The types of data that can be added to a list depends on your platform’s API integration. Typically, they include:

• E-mail address,
• Cookie,
• Card token,
• Phone number.

Rules

The view shows all Rule Sets configured in your application with basic information:

• rule set state: 'active' / 'inactive' / 'simulation',
• the strategy configured for the set,
• run conditions,
• number of active rules in the set,
• shortcut to add a new rule.

You can edit these rule set parameters by clicking the Edit button on the rule set view. It will open the following edit pane:

For adding a new rule set just click 'Add new rule set' button on the bottom of the view.

Adding new rule set

After clicking the 'Add new rule set' button the window for adding a new rule set will open:

Put a name for created rule set, choose an initial state and strategy, add also some rule set run conditions.

You can add more run conditions by clicking the '+ AND' button:

The rule set will be created immediately after clicking 'Save' button.

Rule set states

Rule set states can take the following values:

• Active - rules inside the rule set will be triggered and the outcome will affect the final result;
• Inactive - rules inside the rule set will not be triggered;
• Simulation - rules inside the rule set will be triggered but the outcome will not affect the final result.

Strategies

If at least one rule returns overriding_accept the final recommendation is always accept (all other are overridden) regardless the configured strategy.

Strategy “worst case” - a set returns the worst of rules' recommendations:

• if any rule returned 'refuse' - the set returns 'refuse',
• if any of the rules returned 'review' - the whole set returns 'review',
• if all rules returned 'accept' or 'skipped' - the set returns 'accept'.

Strategy “best case” - a set returns the best of rules' recommendations:

• only if all rules returned 'refuse' - the set returns 'refuse',
• if all rules return 'review' - the whole set returns 'review',
• if at least one of the rules returned 'accept' - the set returns 'accept'.

System gives the final recommendation based on sets' recommendations. Applied strategy is always the 'worst case' strategy (however "overriding_accept" still overrides).

You can see the results of decision engine in the Decision Logic widget in the inquiry view.

Run conditions

Run conditions are simple selectors which indicate when one should run the rule set. If you want to launch different sets of rules for different use cases (e.g. different country, origin), selectors make it possible. They are based on tags assigned to an inquiry and there are two types of the condition configuration:

• "has any" - fulfilled when the inquiry is tagged with ANY of the indicated tags,
• "has none" - fulfilled when the inquiry is tagged with NONE of the indicated tags.

A rule set will be executed when all tags-based conditions will be fulfilled.

Rule set details

By pressing the icon with the number of rules in a specific rule set, you can go directly to the rule set summary.

The view shows a list of rules configured in this set. You can also see the deactivated rules in 'Deactivated rules' tab. Note that these rules will not be triggered.

If you'd like to try out a rule, you can set it to 'Simulation Mode'. The rule icon will turn grey. It will be executed, however, the rule outcome will not be taken into account when calculating the final result.

You may edit a rule by clicking on it or create a new one using 'Add rule' button.

Adding new rule

The first step in rule creation is to select a rule type. Rule types are grouped into two categories: Basic and Advanced.

The basic group contains the most frequently used rule types. Each of those types has a dedicated, user-friendly interface.

The advanced group contains a catalog of all other rules available in the system. Those rules have names in a technical format and the edit interface is based on a simple text editor.

There are 4 types of basic rules:

• Logical Statement Rule
• List-based Rule
• Velocity Rule (depreciated)
• Rule On Signal

Each rule can be set as:

• Active - the rule will be triggered and the outcome will affect the final result;
• Inactive - the rule will not be triggered;
• Simulation - the rule will be triggered but the outcome will not affect the final result.

Please note that the rule set state takes priority over a single rule state.

Logical Statement Rule

Click 'Create Now' to open Logical Statement Rule creator. This powerful tool allows you to build complex expressions in an easy way. You can connect multiple expressions using logical operators: AND and OR with an unlimited level of nesting.

Step 1

Provide the Rule Name and select the state

Step 2

Click 'Configure expression'

Step 3

Each expression consists of:

• Value A
• Operator
• Value B

Value A

The possible elements that you can input under Value A are:

• Attributes - one of attributes that system receives from you through the Inquiry API integration, e.g. Customer / Email
• Velocity - a formula where you can count e.g. the number of inquiries or the transaction amount or distinct BIN number etc., from the inquiries with the same value of a different parameter, e.g. the same IP. It's also mandatory to define the time frame, e.g. in last 24h. Optionally, you can add additional filters, that will include only a part of the inquiries in the formula
• Distance - select 2 attributes that are related to location, System will count the difference between the values of those attributes
• Text similarity - select the parameters to check for the similarity ratio, which is calculated based on Gestalt pattern matching where 0 means no similarity and 1 is full similarity.

Value B

Possible elements that you can input under Value B vary depending on what was selected as Value A, you can choose from:

• Attributes
• User input

Operator

The operators used to in the comparison in the expression are:

• Relational operators: =, !=, <, <=, >, >=

• matches (REGEX) and not matches (REGEX) - this operator will compare the value of the attribute, that was selected in Value A with the input that you will provide as Value B. The input under Value B should keep the format of the regular expression, you can use meta characters:

  • [ ] - matches a single character that is contained within the brackets. For example, "[abc]" matches "a", "b", or "c"
  • ^ - marks the start of the string. For example, "^Com" matches "Computop"
  • $ - marks the start of the string. For example, "utop$" matches "Computop"
  • \ - marks the next character as literal. For example, "\(" matches "("
  • * - matches the preceding character zero or more times. For example, "Compu*op" matches "Compuop" and "Computop"
  • + - matches the preceding character one or more times. For example, "Compu+op" matches "Computop" and "Computtop"
  • ? - matches the preceding character zero or one time. For example, "Compu?op" matches "Compuop" and "Computop"
  • . - matches any single character
  • | - "a|b" matches "a" or "b"

• in and not in - this operator compares the value of the attribute, that was selected in Value A with the various inputs that you will provide as Value B

• is substring and is not substring - this operator checks if the value of the attribute under Value A is a substring of the value of the attribute or user input under Value B

• contains and not contains - this operator checks if the value of the attribute under Value A is a contains the value of the attribute or user input under Value B

Step 4

Once you create the expression, you can add another one. You can connect multiple expressions using logical operators: AND and OR with an unlimited level of nesting, e.g:

Step 5

As a last step, select the recommendation that should be provided in case all the conditions from the rule are met and what should happen otherwise, e.g.:

List-based rule

To configure a List-based Rule you need to follow 3 steps:

Step 1 Select inquiry attributes.

Select one or more attributes that will be searched on the list.

Step 2 Select the list.

Select an existing list or create a new one. The list you choose must have a valid structure for the selected match conditions.

If you decide to create a list, it will automatically assign the same fields as the rule. Confirm list creation by clicking “Add list”. The unique name of the list is validated and the newly-created list is selected. If the name of the list is duplicated, you will be asked to change it.

If you created a new list while creating the rule, the list will be visible in Accept/Refuse lists view.

Step 3 Select outcome.

Choose what the rule should return either on match or miss. Thanks to it, you can create multiple scenarios using only one type of rule.

Rule on Signals

To configure a Rule on Signals, you need to follow 2 steps:

Step 1:  Select one or many profiler signals from the list of all available signals.

Platform-agnostic signals

Web signals

iOS signals

Android signals

Step 2 Define how many signals from the list must be triggered to give a selected recommendation.