AI Interactions: the revolution of the mail channel
75%
Acknowledgment of intent
60%
Reduced service costs
8
Weeks of supervised training
75%
Acknowledgment of intent
60%
Reduced service costs
Challenge
A leading Utility client of ours presented the need to automate the handling of requests received from end customers on the mail channel, which generated time consuming activities on operators who manually handled the requests without adding, in most cases, any value.
Our solution
To address this need, we configured and trained our Intelligent Automated Journey solution to interpret customer intent and communication sentiment, providing for automatic feedbackof feedback to the customer, with predefined responses or routing to self-service channels, or triggering escalation to a dedicated human queue.
EDUCATION
We carried out 8 weeks of training on our solution using a supervised approach, leveraging both the 'universal' knowledge base and that of the client company surveyed from a significant and representative sample of different case histories.
ARTIFICIAL INTELLIGENCEÂ
All emails received are processed by the AI in order to identify the customermain intent of the customer, any key data and the interactionsentiment of the interaction. Based on these informations, each request is categorized, accompanied by immediately understandable summary indicators on the reliability of AI processing, and handled according to the defined path.
SMART AUTOMATION
In addition to the production of the automated responses, for those instances where it was necessary toacquire additional information from the customer, we planned to direct the customer to dedicated and customized web pages for the easy and immediate acquisition of this information.
HUMAN SUPPORT.
Needs categorized withless than 96% reliability or 'negative' sentiment are routed to our Customer Specialists who, concurrently with handling the customer need, track the correct categorization of the email by contributing on going to AI engine training.
