Radiant
A software suite that records and transcribes the calls to the callcenter, manages callcenter staff workload, helps to track down a potential issue
Problem statement
It is not an easy job to be a supervisor of a big call center. His primary task is to ensure that each of the thousands of calls daily helps the clients or is at least correct and polite. In case an issue pops up, he would have to quickly find the call that caused it and the exact place inside this call. He also needs to manage the workload of the call center employees to distribute the load and prevent burnout.
Widgets and app structure
All applications from the suite share the same design system, spatial structure, and high-level widgets like the wave playback control, filtering panel, or sidebar. They all have their constant place in all the apps and that helps to create a consistent user experience.
This app records, transcripts and labels all calls. It helps the auditor to find a particular call, navigate and listen to a certain part of this call and evaluate the overall answer quality of the technical support team.
The second application manages what is currently going on in the call center team. Who is on the call, who is online. What recent calls have been made. This application also gives the user an ability to find a certain place in the certain call.
The third application is a simple call recording tool for the call center personell.
Design evolution: from clickable prototypes in Axure and a conceptual design to the final UI mockup.
How to show the conversation between two or more people in a visual form? How to display where one person’s replica finishes and another person starts speaking? How to navigate through this efficiently? In 2024, that answer seems obvious. Let an AI transcript the converation into text and use it for navigation. But at the time when this project was developing, speech to text technologies were in their infancy and could not provide an output reliable enough. That is why here I introduced this wave control. The UI highlights each person’s replica (this is possible because they are recorded and stored separately). The supervisor can listen to the conversation from any moment in time, read the transcription, and make annotations.