“AI plus emotion intelligence will be the next great thing”

Israel based Amir Liberman is a researcher and an expert in the field of voice analytics and emotion detection. As the CEO of Nemesysco, an Israeli company that provides voice analytics technology and solutions for genuine emotion detection he has been instrumental in developing and expanding a patented Layered Voice Analysis (LVA) technology that reveals and measures the genuine emotions of a speaker during voice based communication. This technology has found significant traction in the Indian market across sectors especially with the Indian security agencies. In an exclusive interaction with THE WEEK he discussed the significance of voice analytics and emotion detection and how important it has become in the present times to catch terrorists, fraudsters and also in the HR recruitment space besides many other fields.


How does voice analytics technology work?

Voice analytics technology primarily focuses on the sub-phonetical elements in the human voice. These are elements of the voice that we can hardly notice with our human ear, but this is an area where traces of our real emotions hide. Voice analytics has become very important across segments today as it can help in detecting excitement, stress, uncertainty, anger, happiness, hesitation, embarrassment and more in the human voice.


How important is voice analytics and human emotion detection in the present scenario?  

This technology has become very critical these days. Security organizations are using voice analytics to perform activities ranging from assisting investigations remotely or on site, build psychological profiles on relevant people of interest, analyze intelligence from various types and assist in ensuring public safety and preventing terror attacks. It is also playing a very important role in the call centre segment also as managers can filter through the mass volumes of calls to identify mishandled processes and improve service levels. Business flaws that cause customers to call service centres in anger or with frustration can be identified instantly with this technology and help to cure the problems from its core. On the other hand it can also help in the HR and the recruitment processes as it can help identify higher quality candidates in a much shorter time frame than existing processes. For instance our LVA technology based on voice analytics has enabled new personality tests, which take less than twenty minutes for a candidate to complete and provide recruiters with several important metrics that can give insight into the candidate’s personality and predict with high levels of accuracy the future behaviour of an employee in ways that were never available before. Emotion detection has also assumed great importance in many fields and we expect to see innovations in its applications especially in the education, healthcare and wellness fields too.


What kind of work have you done in this technology segment?

We had initially developed the LVA technology many years ago for security applications. Globally, increasing the abilities of security organizations and their staff to maintain safety and prevent crime and terror is one of our most important missions. In the corporate world, we seek to serve different vertical applications, helping organizations select more suitable employees, increase service levels and communications with customers and internal business units, and fighting fraud by using our LVA technology. For instance the cost of hiring the wrong candidate for a job can be significant and this technology can help to improve the entire recruiting and onboarding processes from vetting to training and from honesty checks to the personality and work suitability tests. These help not only the employers, but also job seekers to find the right organizations and positions for them, where they can feel most comfortable and meaningful. On the other hand for banks, insurance companies and other financial services providers, fraud costs these businesses enormous sums of money each year. We are working with financial service providers around the world to fight fraud. Broadly speaking our approach is different to traditional approaches to voice analytics, which generally focus on analysing the content of speech, or other alternative approaches that rely on expressed vocal intonations, which can easily be fabricated by a speaker.


What kind of R&D work have you done in this field?

We have a talented R&D team in Israel. Over the past 24 years of the development of our technology and product portfolio, many different people have participated in our research and work, including prominent psychologists, psychiatrists, polygraphists, phoneticians, AI (Artificial Intelligence) experts and domain experts from all our relevant fields of activities. Today, most of the new research is performed in universities by students as part of their PhD work. We are working closely with several universities around the world, including prominent universities in Japan, Russia, Israel, the United States, China, Hungary and India. In India Gujarat University is providing training on our LVA technology and some of our investigation focus tools as part of its criminology studies program.


What is the future scope of this technology in India?

India is becoming an interesting hub of AI technologies and I have recently seen many impressive AI-based research projects that are currently being developed in the country. From my own research into AI, I know that real AI can only be utilized when the technology can understand emotions and factor those into its decision processes when dealing with humans. I think AI plus emotion intelligence will be the next great thing.