Enable a generative AI capability in Virtual Agent Designer
Release version: Washingtondc
Updated February 1, 2024
1 minute to read
Add generative AI functionality to Virtual Agent topics to generate text, summarize information, analyze user sentiment, and interact with large language models (LLMs).
Before you begin
You must be on Vancouver patch 2 or a later release.
Role required: admin
Procedure
Go to the OneExtend Builder Configs table by entering sys_one_extend_builder_config.list in the navigation filter.
Open the Virtual Agent Designer record.
In the OneExtend Builder Capabilities related list, select New.
On the form, fill in the fields.
Table 1. OneExtend Builder Capability form
Field
Description
Capability
Capability that is available in the Virtual Agent Designer with your preferred provider. You must have at least one provider for the capabilities, but you don't need to use the same provider for all the
capabilities.
Execution Mode
Mode that determines when a capability is executed in relation to other processes. For Virtual Agent Designer, select Async. The Sync and Fire and Forget options aren’t supported.
Figure 1. Generic Prompt builder definition form
Select Submit.
Repeat steps 3 through 5 for each capability that you want to enable.
Result
Generative AI capabilities are available for Virtual Agent Designer.
What to do next
Create topics and topic blocks and enhance your Virtual Agent with generative AI capabilities.
You can also access the Flow Designer actions directly with the Action topic block in Virtual Agent Designer.