Artificial Intelligence Online

Improved service quality in call centers due to automatic voice analysis by audEERING


The competence and professionalism of call center service agents is an important precondition for customer satisfaction. Systematically training call center agents requires an automatic analysis of various conversation parameters, like meeting the goals set by the company and defining the predominant atmosphere during the call. This is what the company audEERING, founded by the internationally renowned Affective Computing expert Prof. Björn Schuller (University of Passau and Imperial College London), is specializing on. Together with the Swiss company Spitch, a leading provider of solutions for automatic speech recognition (ASR), voice user interfaces (VOI), and speech analytics, audEERING is integrating new technologies for automatic recognition of emotions in call center environments to allow for personalized training of service agents and improvement of service quality in call centers.

Gilching/Oberpfaffenhofen – Zurich, 28.02.2017. The market for call centers is growing steadily since almost every company is nowadays offering their customers some kind of a service desk. Online shops, mobile phone providers, banks or insurance companies offer instant support by telephone for their customers. But not only the number of call centers is growing – so is the frustration on both ends of the line. Angry customers complain about long waiting times and swamped call center agents – about high levels of stress. To continuously improve service quality, agents need to be trained regularly – not only on technical aspects but also with respect to their general communication skills. The evidence base for such trainings are automatically analyzed calls.

One of the most important parameters for customer satisfaction during a service call is the predominant atmosphere, determined by emotions of agent and caller. Emotions may also affect the customer answers affecting Net Promoter Score (NPS) measurements. Due to decades of research at the Technische Universität München (TUM), the company audEERING, which was founded by former TUM employees, is able to offer an automated solution for emotion and personality recognition. Modern approaches for pattern recognition allow for reliable automatic recognition of emotions from speech signals. The core technology for this comes from the software toolkit “openSMILE” which was developed by audEERING and has received multiple awards.

In a joint project with the Swiss company Spitch, audEERING’s audio analysis methods are being optimized for service call recordings. For calibrating of the recognition software with respect to the service call domain, real call center audio recordings are used and labeled by experts in order to “teach” the software extracting emotions in a learning phase. The labeling of speech recordings is performed by the innovative crowdsourcing technology applied within the iHEARu-PLAY platform, which was developed by audEERING in cooperation with the University of Passau. iHEARu-PLAY is a platform allowing its users to assign emotions to recordings via the internet.

“Every innovative company nowadays explores new approaches for productive customer interaction. Speech, voice, and emotion recognition can be seen as essential elements of a continuative business strategy. Spitch is a leading provider of speech and voice recognition and we are happy about the cooperation with audEERING, the leading vendor in the area of emotion recognition from audio. Together, we offer a complete integration of identification, verification, speech recognition, and emotion recognition, allowing our customers to use this potential for competitive advantages”, says Alexey Popov, CEO and founder of Spitch.

„By this trend-setting solution and cooperation between audEERING and Spitch, companies can not only profit from enhanced customer satisfaction due to better training and prevention measures – they also get support in improving psychological stress of call center agents. This generates added value on all areas”, adds Univ.-Prof. Dr. Björn Schuller, CEO of audEERING.

About audEERING GmbH

audEERING GmbH was founded in 2012 and is a leading company in the area of intelligent audio analysis. Its product portfolio comprises software systems for automatic emotion and speaker state recognition from speech signals and methods for music signal analysis. Furthermore, audEERING offers custom solutions for audio analysis based on its award-winning software toolkit “openSMILE“. Since 2015, audEERING hosts an ERC 2015 proof-of-concept grant for excellence science by the European Research Council in the area of advanced audio signal analysis for emotion and state recognition from speech. The company’s headquarter is located in Gilching near Munich. audEERING’s customers comprise DAX companies in the telecommunications and automotive industry as well as leading market research organizations, national and international media companies and service providers. For further information visit www.audeering.com

About Spitch AG

Spitch is a rapidly growing Swiss based Speech Analytics and Spoken Language Technologies company with a global perspective and a bespoke approach. In addition to Switzerland, Italy and the U.K., Spitch is currently active in the German, U.S., French, and Russian markets. Spitch is the first provider of solutions for Swiss German (Schwiizerdütsch) and its many dialects. For further information visit www.spitch.ch

Contact:
Dr. Martin Wöllmer
audEERING GmbH

Office:
Friedrichshafener Str. 3
82205 Gilching
Germany

mw@audeering.com
+49-8105-7756150
+49-175-5112554

This release was published on openPR.