In the News / 28.08.2019

India AI player Uniphore eyes post-call center world

Fresh from raising a $51 million Series C round, Uniphore’s founder argues that artificial intelligence will become as much about quality of engagement as back-end efficiency

“The consumer preference is now shifting to being omnichannel. The consumer chooses which medium to reach the enterprise, at what time day or night, and the enterprise needs to be equipped to service that customer at that time,” says Umesh Sachdev, co-founder and CEO of Indian enterprise artificial intelligence (AI) developer Uniphore. “No one wants to call call-centers anymore, go through an IVR [interactive voice response], wait in a queue, and then talk to an agent who is neither empowered nor has the full information.”

Sachdev is taking this philosophy to the speech recognition segment, where he says AI that improves customer experience in both routine and complex situations is set to produce a number of unicorns in the $350 billion global market for “conversational service automation.” US investors appear to agree. Uniphore has raised a $51 million Series C round led by Santa Monica-based March Capital Partners with support from Silicon Valley players including as CXO Fund and National Grid Partners.

The plan is to scale up in the deeper B2B markets of the US with technical support from the robust local AI industry. March Capital, for example, runs an incubator in Palo Alto called The Hive that manages an AI-focused fund. One of the GP’s portfolio companies, an AI-powered cybersecurity player called CrowdStrike, went public last month and now commands a market capitalization of almost $20 billion. “All of that knowledge is now available to us,” Sachdev adds.

Uniphore relocated its core operations to California last year to make this happen, although it still maintains an office with 200 employees, mostly engineers, in Chennai. The company claims to have the only platform globally able to listen and act on language, including specific dialects, detect and analyze customer emotion, and then predict true intent, all in real-time. The key to making it successful, however, will be to avoid fully automating and embrace the “human-in-the-loop” approach.

“When the AI already knows me, it tells the agent on the other side who I am, what I did in the past, what my preferences are,” Sachdev explains. “So when I ask a question on something complex like healthcare or types of insurance, instead of a call-center representative putting me on hold and finding the answer, the AI already has it pulled up in front of the agent. It’s a combination of human beings and AI working in tandem that is the real answer – not a hyper-bot that sounds like a human.”

 

Source: AVCJ.com, published: 23 August, 2019