Tip: To train your agents, you can start by enabling recommendations for a single skill.

Use a recommended bot or article

In Conversational Cloud, bot and/or article recommendations are displayed inline within the conversation.

To use a recommended bot

  • Click Use bot to join a bot to the conversation, so the bot takes over. You stay in the conversation, so you can monitor the bot’s progress and remove the bot if needed.

    Tip: As shown in the image above, a system message announces when the bot joins the conversation. You can customize this message by changing the bot user's nickname.

To use a recommended article

  • Click Use Article to copy the article’s text to the agent’s text input area. You can edit the text before sending it to the consumer.

Vote up or down to train the model

To provide feedback on a recommendation, click the Thumbs up button or the Thumbs down button.

“Thumbs up” tells the underlying Conversation Orchestrator model that it was a relevant recommendation, while “thumbs down” tells it that it’s not. The relevance score is calibrated against this feedback so that Conversation Orchestrator can continuously improve its recommendations and provide the most relevant content to agents.