For the complete documentation index, see llms.txt. This page is also available as Markdown.

Data types and data inventory

Introduction

Data type describes the structure of a fragment of the data model, e.g. applicant data, product parameters or property data. After assigning a data type to a model node, the editor suggests field keys together with their descriptions and value types, and the platform recognizes the type of data stored in that fragment of the model.

Data inventory is a list of information stored in the business application model. It is divided into categories, e.g. identification and contact data of persons, finances or properties. Each type belongs to one inventory category, and each of its properties is an item in that category. This makes it possible to know, for each application instance, what kind of data has been saved.

The inventory includes all data described by data types, not just personal data. It also includes product data, e.g. loan parameters.

Data inventory is used, among others, to:

  • indicate to the platform personal data subject to GDPR mechanisms (depersonalization, export, deletion) - see GDPR in the data model,

  • passing saved data to the organization’s data register system in an agreed structure common to all applications,

  • controlling the type of values saved in the data model - see Field value type (typing).

The data type catalog is configured separately for each platform deployment. If the data model editor does not display the field Data type, the deployment does not have a data type catalog configured. The examples on this page concern a mortgage loan application - the catalog in a given deployment may contain different types.

Basic concepts

Concept
Meaning

Data type

Description of the structure of a fragment of the data model: a list of properties with descriptions and value type. It is selected in the Data type node field.

Inventory category

A group of data in the inventory to which the data type belongs, e.g. Person identification data or Properties. The category is determined by the type, not by the node key in the model.

Type property

Field defined in the data type, e.g. firstName or city. It corresponds to an item in the inventory category.

Nested structure

A property that itself has a structure, e.g. an address or a list of contact channels. It may belong to a different inventory category than the parent type, e.g. Person contact data.

Constant value

A property value defined in the data type, e.g. subject type or product code. It is filled in by the platform, so it is not declared in the model.

Out-of-schema field

A node field whose key does not correspond to any type property. It is allowed, but it does not go into the inventory.

Setting a data type on a node

  1. Open the application data model in edit mode.

  2. Click Add field and in the field Key enter the node name, e.g. property.

  3. In the field Data type choose a type from the list, e.g. Property. Next to each type the list shows its description.

Lista typów danych w karcie węzła modelu danych
Figure 1. Selecting a node data type

When setting the type, the following rules apply:

  • The type can be assigned only to a node whose ancestors have no assigned type. Child nodes inherit the type of the ancestor - the field Data type is not displayed for them.

  • Assigning a type does not automatically add fields to the node. Fields must be added separately by selecting them from the type property list.

  • Removing a type requires clearing the field Data type. The node fields remain unchanged.

  • The same type can be assigned to several model nodes, e.g. describing a borrower and a co-borrower. All these nodes are included in the same data inventory category.

Adding fields from a data type

  1. In a node with a set type, click Add field.

  2. Expand field Key and select a type property from the list, e.g. propertyKind. Next to each property the list shows its description and value type. To find a property, type part of its name in the field Search or enter name.

  3. Repeat steps 1-2 for each data item that the application must store in the model.

Lista właściwości typu danych w polu Klucz
Figure 2. Selecting a field key from the type property list

The list contains only properties that have not yet been assigned to other fields of the same node. Selecting a list property, e.g. a list of contact channels, automatically marks the field as Array.

After entering a key that is not on the list, the editor labels the field as Out-of-schema field. The field’s behavior does not change, but it does not belong to the inventory and is not covered by the mechanisms that use it, e.g. personal data handling.

Constant values

In the data type definition, you can specify a constant property value. For example, in the type Applicant the subject type has the value natural person, and in the type Loan parameters a fixed product code is defined. Such a value is identical for every node of that type, therefore:

  • the editor does not suggest this property in the key list,

  • the platform passes the constant value each time the node data is saved, even if the save does not contain any other values of that node,

  • the constant value is not stored in the data model and is not returned when the model is read.

If, nevertheless, a field with the key of a property with a constant value is added to the node, the editor will mark it as Out-of-schema field, and the platform will pass the value resulting from the type - the value saved in the field is ignored.

Data inventory in a running application

Passing inventory information is enabled separately for each platform deployment. If it is not enabled, data types still organize the model in the editor, but the data save does not carry information about inventory categories.

When passing inventory information is enabled, the platform, each time values are saved to the data model:

  1. Determines the nearest node with a set data type for each saved field.

  2. Based on the type and the field path relative to that node, determines the inventory category and position.

  3. Appends this information to each saved value in the data model save event, together with the type’s constant values.

The receiving system for events, e.g. the organization’s data register system or the component performing depersonalization, does not need to know the definition of the application data model - the information contained in the event allows it to determine what data was saved and what category it belongs to. Since the category is determined by the type, not by the key, the same type used under different keys or in different applications ends up in the inventory in the same way.

Example

The application handles a mortgage loan application. The data model contains three nodes with set types:

Node
Data type
Inventory category

borrower

Applicant

Person identification data

mortgage

Loan parameters

Loan parameters

property

Property

Properties

This is what this fragment of the model looks like in the source view (attributes irrelevant to the example omitted):

In this configuration:

  • attribute type appears only on nodes borrower, mortgage and property - their descendants do not have their own type,

  • fields contacts.channel and contacts.value belong to the category Person contact data, because the property contacts of the type Applicant has its own nested structure,

  • the model does not contain fields for the subject type or product code - the platform passes them as constant type values Applicant and Loan parameters.

Where do data types come from

The data type catalog is prepared by the platform implementation team based on the organization’s needs. Data types, their properties and inventory categories are shared by all applications in a given deployment. The data model editor allows you to select them, but does not allow you to create or modify them. Missing required type or property should be reported to the implementation team.

Best practices

  • It is worth assigning data types already at the stage of designing the data model, before forms and scripts start referring to its keys.

  • Field keys should be chosen from the list of type properties rather than typed manually - a field outside the schema does not go into the inventory.

  • Fields intended for constant values defined in the type should not be added to the model - the platform fills them in.

  • Information that can be retrieved dynamically from source systems using data sources.

Last updated

Was this helpful?