Using Decision Designer

Modeling data

Business decisions apply to real-life data such as loans, cars, or customers. In Decision Designer, you describe the data that governs your business activity in a data model. Creating a data model vocabulary allows you to populate decisions in your own language and puts you in control of the business logic.

The data model vocabulary can be easily created by using text fields and drop-down menus to list the specific characteristics and attributes of your business objects. For example, you describe a car that has a model, a category, and a registration year. These attributes have values in the real world, and decisions depend on these values.

The data model vocabulary can be used across the various models within a decision service.

Modeling decisions

Decision Designer provides two options to model complex business decisions: decision models and task models.

Decision models offer a straightforward and low-code approach to express and refine business decisions through a structured, visual representation of a decision. Using this approach, you can model business decisions by specifying:

• The information that is required to make the decision.
• How to make the final decision with that information – the decision logic.

Each decision model contains a decision diagram. Diagrams provide an abstract, high-level representation of how decisions and the data that is required to make these decisions are structured and related to each other. Creating diagrams is an iterative process where you decompose the decision that you want to make into simpler decisions. Decision decomposition helps to make the decision model simpler and easier to manage.

Diagrams are composed of a set of nodes that are used as building blocks to represent decisions in a graphical way:

The drag-and-drop feature available in the modeling environment helps you create and arrange nodes in a logical flow and draw dependencies from node to node.

Task models offer you a more advanced way to define decisions. In a task model, the decision logic is defined at the root of the model, outside of a diagram. It is implemented as sets of business rules and decision tables that can be organized and grouped in folders. Each task model contains at least one ruleflow to control the execution of rules.

While a decision diagram represents dependencies between decisions and input data, a ruleflow defines a sequence of tasks. It chain tasks together, and specifies how, when, and under what conditions they are executed. A ruleflow consists of several task nodes that are connected by logical links:

The ruleflow editor palette helps you create and arrange ruleflow elements. The drag-and-drop feature available in the ruleflow editor allows you to easily draw transitions from element to element.

Enhancing decision-making accuracy

Decision Designer brings together prescriptive rules, machine learning, and artificial intelligence to make business decisions with greater accuracy.

Predictive models

Predictive analytics and machine learning provide valuable insights that can help businesses make more effective decisions. For example, machine learning algorithms can help identify at-risk customers, fraudulent claims, failing machines, competitive orders, and cost-effective suppliers. However, these insights must be acted upon to provide value to the organization. This can be achieved with the help of decision automation.

Decision Designer takes the complexity out of combining predictions and decisions: in predictive models, you can use a step-by-step wizard to prepare remote machine learning models for consumption or import ruleset and scorecard machine learning models and turn them into transparent business rules or decision tables. When a predictive model is complete, you encapsulate the predictive model in a decision node in your decision model. When the decision model is executed, the machine learning model computes a prediction based on the inputs of the decision node that contains it.

Generative AI models

By integrating generative AI in your decision services, you can seamlessly process unstructured data and plain text within your decision models and task models. By incorporating natural language understanding and generation capabilities, your decision services can handle and interpret human language more effectively, enabling more sophisticated and context-aware decision-making.

You can use the prompt editor to analyze or generate specific content based on the context of the decision service itself. You can also provide inputs and expected outputs as examples to improve the performance of the generation with few-shot prompting.

Collaborating on decision services

Collaboration between business users is essential to the success of decision services. In Decision Designer, users who have access to the same decision automation can collaborate on the decision services it contains. When you make a change, it is automatically saved locally in the application and visible to you only. When your work is ready, you can share it with your collaborators.

Working in a shared space not only makes it easier to track changes made by other users, it also simplifies the management of decision automation versions and the resolution of conflicts that are caused by competing changes to a file.

Deploying decision services

When a decision service is ready for deployment, it can be securely deployed directly from Decision Designer before being rolled out to production.

Deployed decision services can be called by any client application by using the decision runtime REST API. The decision runtime offers APIs for executing decisions and managing decision service resources: run specific versions of decisions, manage deployment spaces and decision metadata, troubleshoot executions, and much more.