Example of language understanding and hand-off

This is an advanced scenario where a chatbot tries to answer questions about Intelichat, which are made in natural language. When it can't answer, the dialogue is seamlessly escalated to a human.

Note: this demo requires that someone signs up as an agent, so it must be previously arranged with some Intelichat's representative. The language understanding is executed by connection to wit.ai.

It is important that during the demo the representative shows you how to monitor questions that are not answered by the chatbot, the training of the language understanding system and the creation of content in knowledge base. This corresponds to the complete cycle of knolwedge capturing and distribution.

Click here to chat with the chatbot, then ask your questions about Intelichat.

Despite this is an advanced solution, it is simple to design. See below an image of the dialogue flow necessary for this solution.

For internal use only:
wit.ai: Document_english
Intelichat: js@gmail.com/wit.ai and hand-off