La Mobilière's Customer Service team receives a large number of emails every day without having visibility or quantitative metrics on requests. The result is a fragmentation of internal processes and a consequent loss of efficiency.
The project developed with La Mobilière is a combination of technology, knowledge and collaboration. Classifying such a large number of e-mails on very technical subjects was no trivial challenge, but with the use of modern and innovative artificial intelligence models we managed to achieve a high level of accuracy.
Our goal was to help La Mobilière handle its customers' requests more efficiently, saving time and effort for everyone involved in customer service activities. We created a system that classifies incoming requests, based on the collection of 1 million anonymous emails extracted from La Mobilière's databases. An automated service based on artificial intelligence to help classify incoming requests and emails, categorise them and process them in a more sophisticated way. The first step of the project was a semi-supervised analysis of the entire email history of the last few years. This allowed Indigo.ai to create a structured Knowledge Base to map all the different types of requests.
Anonymous analysed e-mails managed by the classifier.
Knowledge Base areas created using the classifier.
classes of responses that can be automated by our AI technology