The design of Conversagent allows for branching pathways and quite complex interactions.  To begin with, we recommend building your Virtual Human with a single Conversation, either using a template, or from scratch. However, once a Conversation becomes quite large, it can become difficult to maintain.

If you are wanting your Virtual Human to deliver a customised experience, then utilising more than one Conversation is useful.

Additionally, when a consumer is interacting with your Virtual Human, they may stop at any time throughout the Program. Any Conversation they are in the middle of is not completed, and therefore any variables set or updated in this conversation will not be saved to the consumer's profile. This means that they won't appear on any custom reports you might have built.

In the same vein, Alerts are only sent to the server at the conclusion of the Conversation they are raised in. If a Conversation is not completed, Alerts raised in this Conversation will not be shown in the portal. Breaking a long interaction into sections allows you to get data from earlier sections, even if the consumer doesn't make it through to the end of the program.