Configure the AI-powered smart suggestions for authoring

As an administrator, you can configure the Smart Suggestions feature for the authors. The smart suggestion service is secured by Adobe IMS auth-based authentication. Integrate your environment with Adobe’s secure token-based authentication workflows and start using the new smart suggestion feature. The following configuration help you to add the AI configuration tab to folder profile. Once added, you can use the smart suggestions feature in the Web Editor.

Create IMS configurations in Adobe Developer Console

Perform the following steps to create IMS configurations in Adobe Developer Console:

NOTE
If you have already created an OAuth project to configure the microservice-based publishing, you can skip the following steps to create the project.
  1. Launch Adobe Developer Console.

  2. After successfully logging in to Developer Console, you’ll view the Home screen. The Home screen is where you can easily find information and quick links, including top-navigation links to Projects and Downloads.

  3. To create a new empty project, select Create new project from the Quick start links.
    Quick start links {width="550"}
    Create a new project.

  4. Select Add API from the Projects screen. The Add an API screen appears. This screen displays all available APIs, Events, and services for Adobe products and technologies with which you can develop applications.

  5. Select the I/O Management API to add it to your project.
    IO Management API
    Add I/O Management API to your project.

  6. Create a new OAuth credential and save it.

    OAuth credential tile in configure API

    Configure OAuth credential to your API.

  7. In the Projects tab, choose OAuth Server to Server option and then select the newly created credentials.

  8. Select the OAuth Server-to-Server link to view the credential details of your project.

    connected credentials {width="800"}

    Connect to the project to view the credential details.

  9. Return to the Projects tab and select Project overview on the left.

    project overview {width="500"}

    Get started on the new project.

  10. Click the Download button on the top to download the service JSON.

    download json {width="500"}

    Download the JSON service details.

You have configured the OAuth authentication details and downloaded the JSON service details. Keep this file handy as it’s required in the next section.

Add IMS configuration to the environment

Perform the following steps to add IMS configuration to the environment:

  1. Open Experience Manager and then select your program which contains the environment you want to configure.
  2. Switch to the Environments tab.
  3. Select the environment name which you want to configure. This should navigate you to the Environment Information page.
  4. Switch to the Configuration tab.
  5. Update the SERVICE_ACCOUNT_DETAILS JSON field. Ensure you are using the same name and configuration as given in the following screenshot.

ims service account configuration

Add the environment configuration details.

Once you have added the IMS configuration to the environment, perform the following steps to link these properties with AEM Guides using OSGi:

  1. In you cloud manager Git project code, add the below given two files (for file contents, view Appendix).

    • com.adobe.aem.guides.eventing.ImsConfiguratorService.cfg.json
    • com.adobe.fmdita.smartsuggest.service.SmartSuggestConfigurationConsumer.cfg.json
  2. Ensure that the newly added files are getting covered by your filter.xml.

  3. Commit and push your Git changes.

  4. Run the pipeline to apply the changes on the environment.

Once this is done, you should be able to use the smart suggestions feature.

Appendix appendix

File:
com.adobe.aem.guides.eventing.ImsConfiguratorService.cfg.json

Content:

{
 "service.account.details": "$[secret:SERVICE_ACCOUNT_DETAILS]",
}

File: com.adobe.fmdita.smartsuggest.service.SmartSuggestConfigurationConsumer.cfg.json

Content:

{
  "smart.suggestion.flag":true,
  "conref.inline.threshold":0.6,
  "conref.block.threshold":0.7,
  "emerald.url":"https://adobeioruntime.net/apis/543112-smartsuggest/emerald/v1",
  "instance.type":"prod"
}

Smart Suggestions configuration details

Key
Description
Allowed Values
Default Value
smart.suggestion.flag
Controls whether smart suggestions is enabled or not
true/false
false
conref.inline.threshold
Threshold that controls the precision/recall of suggestions fetched for the tag that the user is typing in currently.
Any value from -1.0 to 1.0.
0.6
conref.block.threshold
Threshold that controls precision/recall of suggestions fetched for tags across the entire file.
Any value from -1.0 to 1.0.
0.7
emerald.url
Endpoint for the Emerald vector database
https://adobeioruntime.net/apis/543112-smartsuggest/emerald/v1
https://adobeioruntime.net/apis/543112-smartsuggest/emerald/v1
instance.type
Type of the AEM instance. Make sure this is unique for each AEM instance that the smart suggestions is configured on. A use case would be to test the feature on stage environment with “instance.type” = “stage” while at the same time, the feature is also configured on “prod”.
Any unique key identifying the environment. Only alpha numeric values are allowed. “dev”/“stage”/“prod”/“test1”/“stage2”
“prod”

Once you have configured, the smart suggestions icon is displayed in the right panel of the Web Editor. You can view the list of smart suggestions when you edit your topics. For more details, view AI based smart suggestions for authoring section in the Experience Manager User Guide.

recommendation-more-help
11125c99-e1a1-4369-b5d7-fb3098b9b178