The AI-generated assistant stores all versions of the trained FAQ, just in case.
How is this useful?
Having a log of all trained versions of the AI-generated assistant allows you to try out different trained NLP models with your audience. You can also revert back to a past model based on user feedback received.
How do I load different trained versions of FAQ?
- The current model that is shown on the webchat is displayed on the top right corner.
- The past trained models can be seen under previous NLP models. They are listed in chronological order.
- Click publish to use that version of the NLP model instead.
Once you have trained the model, please refresh the page.
