Application

An Ascend Application allows you to programmatically specify a group of related Components. Applications simplify the management of complex, repeatable patterns in Flows using Custom or Ascend-native Applications.

Leverage pre-built Ascend-native Applications for common data tasks or develop Custom Applications to implement your specific business logic programmatically.

Structure of Applications

Applications follow a "define once, use many" pattern. As shown in the diagram above, Applications create a layer of abstraction between your code and the Components they generate.

They are defined in Python, configured in YAML, and can be instantiated multiple times in your Flows. Ascend-native Applications use code provided by Ascend, while Custom Application definitions are written by users in their Projects.

When an Application is configured, it dynamically generates a set of interconnected Application Components based on the provided parameters. These Components appear in your Flow Graph as a cohesive unit that can be viewed either collapsed (as a single entity) or expanded (showing all internal Components and their connections).

Collapsed Application in the Flow Graph shows as a single Component.

Expanded Application in the Flow Graph shows multiple interconnected Application Components.

Key benefits

Applications deliver powerful advantages:

• Reusability: Define code once and reuse it across Flows, reducing duplication and simplifying maintenance
• Modularity: Package complex logic in self-contained units that can be easily understood and modified
• Standardization: Enforce consistent implementation patterns across data pipelines and teams
• Scalability: Easily scale by reconfiguring Applications as requirements change without significant rework
• Simplified management: Central management through configuration files reduces direct code manipulation
• Encapsulation: Hide implementation details behind a well-defined interface

Application development lifecycle

Applications follow this lifecycle:

1. Identify patterns - Recognize repetitive tasks across your data pipelines that could benefit from consolidation

2. Create parameterized logic - Design a reusable pattern with configurable parameters for template files

3. Define application logic - Implement the pattern as either:

   • An Ascend-native Application (for common patterns)
   • A Custom Application (for specialized requirements)

4. Configure the application - Create a YAML configuration that references your Application definition and specifies templates and parameters

5. Generate components - When the configuration is processed, the Application automatically generates a set of interconnected Application Components

6. Maintain through configuration - Any changes to the generated Components must be made through the Application configuration and templates, as they cannot be edited directly

Use cases

• Reusable data processing patterns: Encapsulate repeatable tasks like data validation, transformations, or aggregations
• Sub-flows within larger pipelines: Break down complex flows into manageable units for better readability
• Business logic encapsulation: Ensure consistent enforcement of business rules across workflows
• Uniform implementations: Create standardized implementations of common patterns that can be reused and customized

Conclusion

Applications provide a powerful mechanism for creating reusable, consistent, and maintainable logic. By separating the definition (the underlying code) from the instantiation (the configuration and Component creation), they enable you to build scalable solutions that can evolve with your business needs while maintaining a clean architecture.

Ready to create your first Application? Check out our how-to guides!