Observational Learning occurs when a digital agent observes, learns, and emulates the actions of their human colleagues. While the digital agent may know the language and meaning of a user's question, they may not be able to answer the question. At this point, a digital agent would escalate to a human employee. However, to ensure the digital agent is well-equipped to answer the question in the future, the digital agent should observe the continuation of the interaction, once it is escalated to a human colleague.
When Amelia is asked a question that she does not know the answer to, she escalates to her human colleagues, but she remains in the discussion as an observer. She learns from observing the interaction through to the solution so that she is able to help users with the same or similar questions at a later time.