A copilot mode, i.e. the AI is partially activated, would be very useful. For example, if you don't want to release the AI completely (because you are still not 100% sure that the answer is correct), the agents/executives can generate the answers with AI and they can evaluate (like/dislike) if the answer seems correct to them. If they like it, it is sent and if they don't like it, it is taught how it should be before being sent. This way you start to evaluate how correct the knowledge base is and also learn how to respond on the fly.
Please authenticate to join the conversation.
In Review
Feature Request
3 months ago

Jesús Farías
Get notified by email when there are changes.
In Review
Feature Request
3 months ago

Jesús Farías
Get notified by email when there are changes.