An AMEinfo exclusive
CSA technology understands, analyzes, and automates voice conversations in real-time, including transcription of calls, in-call agent coaching, automation of after-call work, and more.
Uniphore, a leading Conversational Service Automation (CSA) company, recently released survey results revealing UAE consumer perspectives on the challenges and opportunities for video conversations especially that 56% of them reported they spent significantly more time on video last year than in prior years.
The survey highlighted that over the past 18+ months, UAE consumers experienced both appreciation and frustration for what video conversations have to offer.
However, the survey also served to announce Uniphore’s expansion of its CSA capabilities and offerings, combining AI-based voice and video automation technology.
Commenting on the research findings, Umesh Sachdev, CEO and co-founder of Uniphore said: “Through AI and automation technology, companies and business leaders can create better experiences for customers, pick up on nonverbal cues that they may have missed, and provide insights using data that is decipherable and actionable.”
The same survey results showed 85% of UAE respondents are open to using automation or AI tools to improve video conversations.
To find out more about the move beyond the contact center, AMEinfo conducted an exclusive interview with Ravi Saraogi, Co-founder & President APAC, Uniphore.
Expanding the Uniphore offering
When looking at AI in communications, the Uniphore application has traditionally been focused on the contact center, assisting agents with real-time support, doing post-call reporting, and so on. But the company has been keen on extending this outside of the contact center.
“Uniphore has been focused on disrupting a $500 billion enterprise AI and automation market with unique capabilities that analyze and interpret the customer experience (CX) at the contact center where the majority of conversations take place and derive better engagement and productivity solutions,” Ravi started.
Over the last couple of years, Uniphore has empowered its platform with capabilities around AI and NLP technology, but it complimented this with major acquisitions.
In October 2020, the company acquired a third-party license for Robotic Process Automation (RPA) technology from NTT DATA to automate the front office for modern contact centers.
In January 2021, Uniphore also acquired Emotions Research lab, a Spain-based AI company that analyses human behavior through facial recognition of emotions, and combined it with its CSA solution to enhance its offerings.
In July 2021, Uniphore announced its agreement to acquire Jacada, a CX pioneer having the industry’s leading low-code no-code platform for contact center automation allowing to solve complex contact center challenges and empowering the design and implementation of transformational CX quickly and easily.
“We can now offer a comprehensive platform where business users can realize state-of-the-art digital transformation via robust AI-led CX solutions, resulting in significant ROI,” Ravi assured.
How reliable is AI in CX?
Ravi agrees that it is very difficult for machines to depict 100% what a human can do, what they are fthinking or feeling.
“The idea is to bring productivity into play. Even if the system is 70-80% accurate, you can bring about 70-80% productivity improvement in the contact center or in conversations that organizations are having, thus generating tremendous business gains,” Ravi offered.
“We don’t have lie detectors, but, for corporate entities, there are particular ways to check if the call went according to script or if the delivery was as anticipated by the consumer. Similarly, education organizations can verify if school children were alert enough, if teachers put enough markers in place to create attentiveness or asked the right questions or not, and if students were responsive.”
The Uniphore system puts enough parameters in place which AI tracks while providing real-time guidance to users be it to improve sales or enhance educational performances.
Uniphore’s conversational automation platform
Conversation service automation using AI has multiple different aspects, including speech recognition to understand what the person is saying, how, and why.
“Uniphore uses voice biometrics to authenticate the person using his voice, detect and understand his/her emotions by the way they speak, and use NLP to understand the intent of the caller,” Ravi explained.
Uniphore Acquisitions that extend this functionality
RPA Technology from NTT Data automates processes in the contact center. For example, if the customer wants to change their address, the bot can do this in the background, allowing the agent to focus more intently on the softer ‘human’ side of the call.
Emotion Research Lab uses AI and ML to identify emotion and engagement levels in real-time over video-based interactions.
Jacada uses low-code, no-code technology that now integrates with the RPA capabilities so enterprise customers can quickly and easily create workflows in the contact center to streamline and automate processes.
“Using bots, Uniphore has unique capabilities to automate certain actions that agents or speakers on the phone or video call need not take anymore. This allows quicker tech deployment for enterprise customers but also with partners across the ecosystem who can now launch a low-code no-code platform across different use cases,” Ravi explained.
“Using virtual agents, customers need not wait a long time to get answers for their queries or for someone to show up on a call. Repetitive responses like verifying fund transfers or address changes are done by our system which is interpreting the right intent, and ensuring accurate delivery is driven across all the systems and elements.”
B2B use cases
When multiple individuals are on a call, it’s hard for humans to keep track of everyone. Uniphore uses AI to help read facial queues and judge interests and address sentiments. This can then be used to generate alerts to participants or tips, during the call, to help better engage with their audiences.
It can have an immediate impact on generating better sales for retail-based organizations as well as creating more effective communications and levels of engagements in audio formats or over video.
In educational institutions, teachers can verify if students are being attentive. With remote work video conferencing, be it with staff or clients, AI can help keep engagement levels between stakeholders high.
Finally, AI can verify if the participants have discussed points meant to be discussed, or tabulate if all the i’s were dotted and the T’s were crossed.
“Gauging sentiment and giving coverage of that to areas around marketing, sales, and internal company collaboration are areas we cover globally,” Ravi said.
“In the region, we have clients in contact centers using our technology, but we are at a stage where we just acquired video AI capabilities, are integrating it in our platform, and working out use case scenarios that can create impact across enterprises in this region.”
Can privacy issues arise from this technology?
It is not unthinkable to assume that when a conversation is taking place, with AI being present to track words said and actions taken, the caller might have concerns about privacy issues.
“When we started around contact center practices, the intent was to keep the agent as a focal point. All the technologies were created around the agents to really uplift their productivity. For video communication, we are not trying to create a sense that someone is watching, but rather uplift the existing communication and bring about benefits to all stakeholders on the call,” Ravi explained.
Ravi went on to say that different countries have different laws around privacy. “Before we start a conversation, we take the caller’s consent by announcing that the call is being recorded, and we ask ‘Are you ok with it?’,” Ravi said.
“What’s more important is that Uniphore is not recording conversations but rather streaming live ones while giving real-time guidance.” It’s only in the event organizations need post conversation analysis that a recording is initiated by consent.