A European-based, peer-to-peer lending platform provider was looking for a solution that would automate customer service operations and enable more efficient processing of incoming messages and queries from different channels.
We deployed our software platform that utilizes the latest technologies in natural language processing. The platform classifies received messages into predefined categories with more than 90% accuracy. Then finds key information in the report like date, name, product name, code, etc. Based on the first two phases, the platform compiles the recommended answers and, if necessary, performs the next automatic step associated with the request contained in the message (creation of a service ticket, change of personal data…).
the platform categorizes more than 80% of the messages correctly from day one and provides ready-to-send answers for 75% of them. This resulted in 60% of all incoming messages being solved with minimal human input. Now it takes only 10 seconds on average to complete the basic customer query.
Let’s stay in touch