Skip to main content

Case studies

Automated e-mail processing

Want instant response to e-mail in the customer service? Artificial intelligence to accelerate communication and increase customer satisfaction. Analysed messages can be used as structured data for further analysis.

One of the world's leading telecoms operators, providing a full range of telecoms and IT services to more than 470 million customers in 150 countries on five continents.


In a competitive market, complying regulatory obligations while ensuring quality, speed, accuracy and control of customer service is an increasingly complex task. The company is always present in the domestic market with innovative products, services to ensure maximum customer satisfaction. With customers at the heart of their strategy, they always strive to exceed their expectations and shape the market ahead of their competitors. The number and complexity of requests through electronic channels is constantly increasing, but the aim remains to respond to all requests as quickly as possible, even within 24 hours, far from the 30 days required by law.


Based on the results of automated analysis, the company can speed up and make the handling of incoming requests more efficient. They can shorten the time to respond by ensuring that e-mails are immediately assigned to the right person according to their priority and subject. Thanks to its flexible design, the system can be easily and quickly integrated into existing analytical tools, allowing analysts to use the results in other areas (e.g. cross-sell, churn) and to perform ad-hoc text analytics.


Implementation of a solution - Emilia – which combines rule-based and artificial intelligence algorithms and can automatically process, categorise emails by topic and business area. The main milestones of the implementation are:

  • Understanding the current business logic
  • Establish topic categories and assign them to subject areas that fit the business logic
  • Processing past email messages to create enterprise-tailored text analytics dictionaries and models that can consistently support the signing of emails to the relevant managers  
  • Integration with existing e-mail handling and analytics systems
  • Automatic generation of reports from incoming messages, allowing operators to prioritise more easily by both urgency and customer type.