Voizz

Comprehensive System for Call Center Efficiency

Comment

Inspiration

The most frequent method we communicate ourselves is through our emotions. Digital assistants like Siri, Cortona, Alexa, and Google Assistant are just a few examples of devices that can perform emotion recognition from voice. When connecting with others, emotions are crucial. Every business that relies on its products unquestionably has a department set out for the customer service center industry. Improved hospitality is essential to the business's expansion. These individuals behave in public as though they were company ambassadors. A call center is a specific industry that caters to other businesses' customer support needs. Monitoring the emotional quotient of customer service staff is necessary for the optimum performance. Our project's primary objective is to determine an employee's emotional state in a call center utilizing that employee's call recording.

What it does

Voizz is a cross platform application evaluate the performance of the call center employees. The performance can be analyzed by following these steps.

  1. Call Recording : Recording the call center employee and customer’s interaction
  2. Audio File Upload : Uploading the recorded audio file in the application
  3. Analyzing the audio file : The audio file is analyzed using speech recognition approach, and employee’s emotional status is obtained
  4. Providing In-app currency : Based on the analysis, the right amount of in-app currency is provided to the employees

Technology Used

React Native, Flask, MongoDB, Librosa, MLP Classifier and Courier APIs.

Courier APIs Used

1. Email
We used the Courier Email API to send the login credentials to staff and for login intimation of staff.
2. SMS
To inform the staff of the in-app wallet transaction, we used the SMS API for notification.
3. WhatsApp
For official business communication, we have additionally integrated the Whatsapp API into our application.

Challenges we ran into

Audio can be exceedingly challenging to categorize using only raw audio samples. Thus, it was difficult to easily classify the audio by extracting its properties. The Librosa package was quite helpful in obtaining these functionalities.

What we learned

We developed the skills and information necessary to deal with a variety of situations and challenges. We also explored the courier API's operation and the incredible applications the courier platform may offer in addition to learning how Courier may enhance interaction and communication between diverse services.

What's next for Voizz

For now, we have created a application that lets us recognise a worker's emotions based on a recorded audio file that the company uploads. As the project is easily scalable, it may be upgraded as and when necessary in the near future. We plan to integrate it as a feature in the customer support department that enables real-time emotion recognition and rewards in-app currency in line with it.

#DevCrew

Built With

Share this project:

Updates