Data modeling basics

• Identify data elements
  • Determine the attributes required for a business process or use case.
  • Align attributes with business requirements for completeness and relevance.
• Refine and normalize data
  • Remove redundancies and inconsistencies.
  • Apply normalization principles to maintain integrity and optimize storage.
• Organize data and define relationships
  • Group data into entities and specify their attributes.
  • Define relationships such as one-to-many or many-to-many to represent real-world interactions.
• Define persistence mechanisms
  • Specify how to store, retrieve, and maintain data over time.
  • Consider database type, indexing strategies, and performance optimization.

Data modeling provides a foundation for application development because it:

• Clarifies requirements: Captures all necessary data elements before development begins.
• Enables process automation: Creates a structured framework for automated workflows.
• Facilitates integration: Establishes a consistent structure for interoperability with other systems.
• Improves scalability and maintenance: Reduces complexity and supports future enhancements.

Effective data modeling is essential for building scalable, secure, and integrated Pega applications. The following points highlight why data modeling plays a critical role in application development. 

• Efficient data management
  Organize and maintain data to reduce redundancy and improve consistency.

• Effective process automation
  Ensure workflows operate smoothly by making properly structured data available.

• Seamless integration
  Support interoperability with systems and objects in the application ecosystem.

• Faster application development
  Reduce complexity and accelerate development with a well-defined data framework.

• Improved data security and compliance
  Support adherence to security standards and regulatory requirements through structured data handling.

• Enhanced reporting and analytics
  Deliver accurate insights by maintaining data integrity and consistency.

• Reusability and maintainability
  Promote modular design for easier reuse and maintenance.

• Scalability and extension
  Provide a flexible foundation that supports growth and future enhancements.

You apply the principles of object-oriented programming, including encapsulation, inheritance, and polymorphism, in enterprise data modeling. 
 
Encapsulation defines the scope and applicability of attributes, determining how developers define each Data Type in a Pega Platform™ application. All data elements that logically fit into the same scope and applicability are grouped to form a data type. 

For example, Street, City, State, and postal code comprise an Address Data Type.

Inheritance defines the reusability and hierarchy, which determines how to arrange each Data Type from most general to specific. 

For example, Data-Party is the most general Data Type for Case participants, while Data-Party-Person is specific to the Data Object of a person.

Polymorphism defines the different forms of abstract objects, which determines how an object can take the required form according to need in the context or hierarchy of the abstract object. 

For example, Data-Bird can be defined as an abstract Data Type. Then, at run time, it can be used as object classes such as Data-Bird-Peacock and Data-Bird-Sparrow.

In Pega Platform, you can use data relationships to model advanced and dynamic data structures. The Data Models are flexible and powerful, supporting concepts such as polymorphism. At design time, you declare a data relationship Field Type that maps to an abstract class. At run time, the required concrete class name is displayed in pxObjClass. 

Business scenario

An auto insurance company's application displays a list of vehicles to cover as part of a quote. The list can include motorcycles, cars, and trucks, each with unique business rules and processes. The following data models are possible solutions for this business problem: 

Solution 1: Create separate data relationships (multiple records) for each type of vehicle

Create a separate data relationship with multiple records for cars, trucks, and other vehicles, such as motorcycles. Each list for the data relationship has a static page class. Developers may need to create separate user interfaces for each page class. 

Solution 2: Create a single data relationship (multiple records) and a single-page class for all vehicle types

An alternative approach is to utilize a single-page class for all vehicle types and a single data relationship (multiple records). To differentiate between processes and rules, conditional logic or circumstancing can be employed. 

Solution 3: Create a single data relationship (multiple records) and a separate page class for different types of vehicles

You can create a single data relationship (multiple records) of covered vehicles where each page is a different class type. Rule resolution uses the run-time class of each page to apply the correct rules, processes, and user interface. 

The following figure shows the pictorial representation of the previous three solutions:

Recommendation

• Solution 3 is the best option. You can add a new vehicle type, and then map the page in the vehicle list to the new vehicle class.
• Solution 1 is not a recommended option because it is not scalable. For example, suppose the application updates to include a new vehicle type, such as Boats. Having multiple page lists might require modification of multiple rules to implement this change.
• Solution 2 is not a recommended option. With only a single-page list class, business rules become more difficult to maintain because there are too many circumstances and more variants of business logic.

The following figure shows the small portion of the Pega clipboard view of the Van, Car, and Truck specializations of the MDC-Data-Vehicle class being added to the same embedded page list, VehicleList:

Pega GenAI Autopilot™ in Pega Platform acts as an AI-driven assistant in App Studio’s Data Designer. It accelerates data modeling by automating object creation, suggesting fields, and simplifying integrations while following best practices. The core functions of Autopilot are:

Automated data object creation
Creates local or external data objects with meaningful names based on application context such as Case Types, Stages, and Steps.

Intelligent field suggestions
Suggests relevant fields for each data object using business logic and design standards to reduce manual effort.

Integration setup
Generates integration details such as endpoints, parameters, and responses using OpenAPI specifications to reduce manual configuration.

Support for complex data types
Handles embedded structures and references to ensure robust and scalable data models.

Conversational guidance
Provides prompt-based assistance for building and refining data models with step-by-step guidance.

Benefits for data modeling

• Speed: Rapid creation of objects and integrations
• Consistency: AI-driven suggestions aligned with best practices
• Accuracy: Reduces errors in data structure and integration setup
• Productivity: Enables developers to focus on architecture and optimization

Best practices

• Validate Autopilot suggestions before implementation
• Use Autopilot for initial setup, then customize for business needs
• Document changes for future reference and governance
   

Check your knowledge with the following interaction: