A large amount of manual effort today goes into supporting customers, be it through voice, chat, or web interfaces. A lot of work has been done to handle many interactions without the need of a human at the other end. Early attempts at self-service have left many users feeling frustrated and ignored. Using AI capabilities we can provide much richer interactions with end users. A well designed dialog management system combined with deep question answering capability is key to a rich interaction. The system can answer naturally phrased questions and find the specific snippet of information that the user is looking for. Users can ask questions in whole sentences and we reply with just what the information that the user is looking for. The accuracy such a system when applied to the scope of information specific to a given task can be quite high, making it a very practical alternative.
This capability can be applied to contact centers in essentially two ways – either to improve the effectiveness of the call center representative or to provide a direct channel for the user, eliminating the representative in the middle. When applied as an aid to the call center representative, the system can be configured to monitor a dialog in the background and suggest the right answers to the call center rep. Alternatively, users can be provided with a completely automated interface to enter their free form questions and get the answers directly from the QA system.