Explore Options for Decision Logic
You can model decision logic using decisions, agentic AI, and integrations. Learn when to use each technology to build decision logic.
When to Use a Decision
When you model decision logic using a decision, you can deliver predictable, repeatable outcomes. Use a decision when all the following statements are true:
-
The decision logic is complex, or it might require updates in the future.
-
The outcome of the decision logic must be the same every time.
See Model Decision Logic.
When to Use Agentic AI and a Knowledge Base
When you upload corporate policies to a knowledge base and allow agentic AI to use the documents to make decisions, you bring human-like reasoning and creativity to your decisions. Use a agentic AI and a knowledge base when all the following statements are true:
-
The decision logic is complex.
-
The corporate policies are likely to change frequently.
-
The decision logic must consider real-time data and current circumstances.
This approach offers several benefits over using a decision:
-
No need for decision logic: Just write your policies as you normally do in a document.
-
Easier updates: All you need to do is update your policy document and then upload the document to the knowledge base.
When to Use an Integration
You can include simple decision logic using a switch action in an integration. Use this approach when either of the following statements are true:
-
The decision logic is simple, and it won't require updates in the future.
-
The logic is technical only, such as enforcing a character limit.
See Route Expressions with Switch Action Branches in Using Integrations in Oracle Integration 3.
