Here is a walk through of how to configure Coveo relevance generative answering. There are four steps. Create the NL models, associate the models to a new query pipeline, create a search page and enable the feature, and finally testing it out. Let's start by navigating to the machine learning section to create the two machine learning models required creating a Coveo RGA experience. I'll first create an RGA model by simply telling the model what content to learn from. Giving it a name and hitting create model. I'll then create a semantic encoder model in the same way. Then I'll head to the query pipeline section to create a new query pipeline. I'll associate both of my models to that query pipeline. Finally, I'll navigate to the search pages section to create a new search page. Here, I'll open the settings section and check the relevance generative answering option to enable the feature in my page. Now I can ask a question in my search page and see Coveo RGA in action. This includes the generated answer, a thumbs up thumbs down option, a copy to clipboard button, a toggle to hide the feature, the citations used to ground the generated answer, and options to reformulate.