Our insurance client had outgrown its own knowledge system which was designed years prior. The system had so much information that it was no longer easy to use and caused customer question sessions to be extended or required handing off.
Ingestion through Alchemy in order to classify and analyse the information.
NLC and elastic search elements on either end of the system.
Voice-to-text to assist agents in data entry and information retrieval.
Form filling implemented into web chat to complete requested transactions.
The system demonstrated that it was capable of learning the information held within the knowledge bank.
It interpreted the questions from the customers and it identified the information that had relevance to the question being asked.
It allowed for a uniformed approach to be established and reduced the use of out of date information or previous variations of policy