Beeline neural network will improve the work of call centers and point out errors to employees

Vimpelcom specialists are developing a training platform that will use speech analytics to evaluate the work of call center employees.

In particular, it will analyze errors and improve the customer service process. With the help of a model that will be built into work systems, an employee will be able to learn even in the process of work: after each conversation, the platform will give him an assessment, highlight gaps and errors, and tell you what information needs to be studied in order to avoid mistakes in the future.

Now we are actively developing a service that transcribes audio into text, and then analyzes its content. The solution will allow employees to hone their communication skills on a voice simulator, improve the quality of customer service through training.

— Big Data Products and Technologies Director Konstantin Romanov

Also, a special team implements a skill testing block in the “mystery shopper” format.

The employee will be called by a robot moving according to a given scenario, the quality of which must be checked. Depending on the answers of the employee, a dialogue develops, and as a result, he receives the test results.