Here, we can discuss Oracle APEX AI configuration, creating an application, using App Builder, and setting up dynamic actions with this AI.
Key Points:
● Configure APEX Assistant Generative AI Services
● Create Apps using Natural Language
● APEX AI Assistant
● AI-Assisted Debugging
Configure APEX Assistant Generative AI Services
Step 1:
➜ Open your workspace > App Builder > Workspace Utilities

Step 2:
➜ Workspace Utilities > Click on Generative AI

Step 3:
➜ Generative AI > Create a Gen AI

Step 4:
➜ AI Provider will ask for different options.
➜ Let we choose the Cohere API which is an open source

Step 5:
➜ If you have an OCI (Oracle Cloud Infrastructure) account,
➜ it's recommended to use the OCI provider.
➜ However, there are certain limitations when opting for OCI, as it supports only the following regions:

Step 6:
➜ Regarding Open AI API it is a paid source.
➜ So I’ll start with cohere API
➜ Provide the name and static ID and enable the app builder to use this AI in application page level.

Step 7:
➜ Generate a new credential
➜ Provide the API key to set up the AI service

Step 8:
➜ After being created, App Builder > Workspace Utilities > web credential has been automatically generated for the AI assistant.

AI ASSISTED APP DEVELPOMENT
Step 1:
➜ While creating an application you can be able to see the new option as Create App Using Generative AI.

Step 2:
➜ For the first time it will ask for the confirmation to access the AI assistant.

Step 3:
➜ Create an application based on your requirement Example: create an application for travel request it should be based on the person travelling from one place to another.

Step 4:
➜ The application has been created with some sample pages. If you are okay with it, we can proceed further. Alternatively, we can modify the application as needed.

Step 5:
➜ The application has been created.

IMPLEMENTING AT THE PAGE LEVEL
APEX Assistant:
➜ If you encounter any issues while using the query builder, you can utilize the "Help Me Fix" feature for assistance.

➜ It will provide with 3 different options
● Use Selection
● Improve Selection
● Explain Selection

Step 1:
➜ Create a page and add a region with the type set to CARD.

Step 2:
➜ Modify the query based on your requirement.

Step 3:
➜ Create an action button and link the project ID and project name to the corresponding AI page.

Step 4:
➜ I have Created a New blank page
➜ I have set up a page item to hold my project ID, project name, and project details.
➜ Additionally, I created a region to show the default content.
➜ To populate the project details, I implemented a Computation that retrieves the necessary values.

Step 5:
➜ In the computation, I have retrieved only the necessary values based on my requirements.
➜ These values can be adjusted as needed upon request.

Step 6:
➜ Create a Dynamic Action in Page load To access the Gen AI.

Step 7:
➜ We have an option called Open AI Assistant.
➜ Select the appropriate service that was previously created.
➜ Here, you can set the welcome message.

Step 8:
➜ We need to set the display as inline and use a container selector similar to a static ID.
➜ This same name should be mapped to the Region's static ID.


Step 9:
➜ The dashboard has been updated according to our preferences.
➜ Now, we can access the AI chat feature.

➜ When you access the Chat AI, you can ask questions related to your project, as shown in the example below.
➜ However, please note that it can only access the mapped data values in the backend.
➜ We need to restrict all the Amount values since we are using the Cohere API, which is open source.

Keep learning.
let's all learn together and make APEX even better.