Invoking Generative AI Use Cases from Siebel UI
To invoke Generative AI use cases from the Siebel UI, create Business Component User Prop as shown below:
Note: User Prop “provider” and “model” are mandatory.
Here's an example for the Sentiment Analysis use case from the Action Business Component for OCI Generative AI endpoints.
| BU User Property Name | BC User Property Value |
|---|---|
| GenerateChatText_Param_1 | compartmentId:ocid1.tenancy.oc1..XXXXXXXXXXX |
| GenerateChatText_Param_2 | servingMode:ON_DEMAND |
| GenerateChatText_Param_3 | input:FIELD=Description |
| GenerateChatText_Param_4 | prompt:SentimentPrompt |
| GenerateChatText_Param_5 | output:FIELD=Sentiment |
| GenerateChatText_Param_6 | lovInfo:LOV_TYPE=GENAI_SENTIMENT;LOV_LIC=POSITIVE,NEGATIVE,NEUTRAL |
| GenerateChatText_Param_7 | outputToOtherBC:BUSCOMP=Service Request,FIELD=Sentiment |
| GenerateChatText_Param_8* | provider:oci** |
| GenerateChatText_Param_9* | cohere.command-r-XXXX** |
Here's an example for the Sentiment Analysis use case from the Service Request Business Component for sample Model Provider.
| BU User Property Name | BC User Property Value |
|---|---|
| GenerateChatText_Param_1 | compartmentId:ocid1.tenancy.oc1..XXXXXXX |
| GenerateChatText_Param_2 | servingMode:ON_DEMAND |
| GenerateChatText_Param_3 | input:FIELD=Description |
| GenerateChatText_Param_4 | prompt:SentimentPrompt |
| GenerateChatText_Param_5 | output:FIELD=Sentiment |
| GenerateChatText_Param_6 | lovInfo:LOV_TYPE=GENAI_SENTIMENT;LOV_LIC=POSITIVE,NEGATIVE,NEUTRAL |
| GenerateChatText_Param_7* | provider:openai** |
| GenerateChatText_Param_8* | model:gpt-4o** |
*New parameters to be added (remaining are predefined parameters)
**Specify value based on "provider" and "model"