CAI technology uses machine learning and natural language processing (NLP) to interpret text, voice, and images to find the right answer. It does that based on patterns within the content we provide it with, processing the semantics of this content to identify the user intents. Through interaction with users the technology learns by itself, becoming more intelligent and improving its responses as time goes by. At the outset the technology needs humans to train it with structured content in the form of question and answer pairs.
Collaboration between developers, UX designers, and information specialists from the beginning is therefore key. You need to get together to determine the business need, define all possible conversation flows, and ensure the technical feasibility. You need to know who your users are, what their needs (intents) are, and how they are most likely to express those needs. Information specialists often know the end users and their needs better than anyone else in the organization, and more times than not the answers to the users’ questions will be contained in the structured content produced by those information specialists. By using metadata and a taxonomy, the rights pieces of information can be provided through CAI to the user.
Pierre-Edouard and Marianne will go into more detail on the synergy between technology and CAI, what is possible now, what kind of challenges we are facing, and the role of information specialists in CAI going forward.