A first audit of the reasons for incoming contacts is made via the semantic analysis of all channels combined (email, phone, chat …). This allows for the identification of the main themes of the most frequent requests and the pain points to be treated first. It then becomes easy to quickly define those for which an automatic response by a virtual agent can be created.


Two virtual agents are implemented to automate answers to the simplest questions, freeing up time for teams to focus on high added-value topics and prioritize tasks & risks. Incoming emails are automatically ranked and prioritized according to their level of importance. A “customer support” chatbot can be set up to answer the most frequently asked questions and guide customers towards the services that can best advise them.

improved customer satisfaction
lower customer support costs