Live Debugging
Integrated debugging is the core part of our authoring platform. You can step through the models and look at the values and parameters. Set breakpoints to suspend execution, inspect the execution context, change values on the fly, understand the execution path, and just about anything else you need to check under the hood.
Breakpoints
Breakpoints allow you to put a Pin on a specific part of a model to pause the execution temporarily at a certain point and control the execution. Once the execution is paused, you are in control to achieve your tasks with full visibility.
On-the-fly change
Change always has a risk. The built-in test framework allows you to build multiple test scenarios without any code based on input and expected outcomes for business rules, decisions, machine learning, and any other models to ensure they do not degrade or regress over time.
Quality Assurance
Pass input and output parameters for simulation, debugging, and analysis to allow step-by-step live interaction with models. This will enable your business to track and measure decision results based on KPIs.
Full Control in Debugging and Simulation
Control in Debug and Simulation
Our advanced authoring platform provides a full set of debugging capabilities allowing you to fully inspect and understand the execution of the logic (business rules, data flows, business decisions…) visually.
- Debug and Simulation: You can provide any input values and data to the authoring platform and put a model into a live and step-by-step debug and simulation mode.
- Navigation Control: In the debug and simulation, you can control navigation to various parts of the model step by step. You can step-into, or step-over and skip based on your requirements.
- Terminate or Finalize: once in the debug and simulation mode, you can terminate the execution or choose to execute the rest of the logic step by step or complete and finalize the execution!
Breakpoints
Breakpoints allow you to put a Pin on a specific part of a model to pause the execution temporarily at a certain point and control the execution. Once the execution is paused, you are in control to achieve your tasks with full visibility.
You can then:
- Navigate to the next step
- Inspect values inside the current context
- Change values on the fly to experiment with a new behavior
- Send the execution context to the Interactive Shell for exploration
- Use the Data Viewer and Boolket to inspect and visualize data
You can manage Breakpoints using the Breakpoints Window by adding and removing them, or temporarily disabling them. You can add Breakpoints to:
- Orchestration steps e.g. Flow, workflow, decision graphs and etc.
- Inside business rules e.g. Decision Table, Natural Language and etc.
Full Visibility
In the debug and simulation mode, you have a full visibility into all the values and data at any step. The Parameters Window and Watch Window will enable you to see all the Input, Output and Local parameters and values.
In the watch Window you can start writing custom expressions to evaluate and prepare and change values on the fly.
Notification and Messages
All the different notification types (Information, Warning, Error and etc.) and messages coming from rules and other parts of the logics are displayed so there is no confusion as what is going to be shown to the end user.
The messages and notification can be localized based on different languages (multilingual messages) and parameterized based on values and special criteria.
Execution Log – Explainable and Traceable
As part of execution, either in debug and simulation or in production all the reasoning behind the made decisions are captures.
So, you can tell exactly why a decision has been made under what circumstances. A full expandability log is provided with special invocation id so it can be associated with a client request for execution.
Easy To Test
FlexRule Designer, our authoring platform has built-in testing framework. This allows you to build unit tests, regression tests and prepare your tests cases to be executed against data input and assert on the expected outcomes. You can build them by drag-and-drop manually or import them in bulk from Excel and test cases will be created for you automatically.
The tests can be integrated into automated CICD pipeline or executed on the manual basis inside the Designer whenever needed to produce reports and ensure quality.
Composite AI model for
Decision Intelligence
