Engage AI

The science

Engage AI is based on many years of research led by Björn Schuller, Chief Scientific Officer at audEERING. With more than 45.000 citations, he is the most cited scientist globally in audio-based affective computing and a top 25-researcher globally within AI (read more here).

A team of AI- and data-scientists, led by audEERING CTO Florian Eyben, and working in conjunction with Jabra’s contact center experts, have optimized and customized this science for the contact center environment.

To get an idea of what the AI delivers, imagine a group of particularly emotionally intelligent people listening to every call, and evaluating the customer sentiment and agent tone.

In the below video, you can hear Björn Schuller and Florian Eyben describe the science in Engage AI:

One very important focus area in Engage AI has been to eliminate biases in the AI-system. Unfortunately, even some of the very best AI systems have been found to exhibit biases, for example towards gender or ethnicity (see for instance this research from Stanford).

Bias can always be present in AI-systems. With Engage AI we take the extra step to ensure that biases are neutralized as fast as they’re discovered. All models are subjected to a large battery of bias-tests before being deployed. And for extra peace of mind, agents can always remove calls from the analysis if they feel they were unfairly scored. If these situations occur, the Engage AI team will work fast to ensure that the AI learns to avoid this bias. After all, continuous learning is exactly what AI is about.

What you like to learn more about the science? Get in touch.