Is it possible to build a voice assistant for a call center that can take turns asking questions to determine the best time to hand over calls to a human? That depends on how high the bar set is. Generally, customers do not wait eagerly. Any Automated experiences on the phone mean voice assistants need to go the extra mile to build goodwill. A 2019 study found that 88% of people would rather talk to a live service agent than navigate an automated system.
If the call center software market is worth tens of billions, there’s a strong incentive to crack the code — or at least get close. Countless tech giants and startups have pitched their technology at the automated call handling problem, from Google (and Google’s Area 120 incubator), Microsoft and Amazon to Got It AI, Replicant and Tenyx. Now, a new startup called PolyAI says it’s working effectively — and points to it as proof. Originally a spin-off from the University of Cambridge, PolyAI provides a voice assistant service that answers calls for companies including FedEx and Marriott.
In a show of confidence from investors, PolyAI today closed a $40 million Series B funding round led by Georgian with Twilio Ventures, Hosla Ventures, Point72 Ventures and Amadeus Capital Partners. CEO Nikola Mrkšić told TechCrunch that the capital, which will bring PolyAi’s total revenue to $70 million and value the company at approximately $300 million post-money, will be used to support the company’s go-to-market efforts as PolyAi grows customer acquisitions.
PolyII’s founders — Markšić, Tung-Hsien Wen and Pei-Hao Su — met at Cambridge’s speech systems group, Mirkić explained, a research lab focused on speech systems. His doctoral work and tenure at Apple (Mrkšić), Google (Wen) and Facebook (Su) laid the foundation for many of the company’s conversational AI systems.
Image Credits: PolyAI
“Enterprises are deploying our technology at an accelerated pace due to unprecedented labor shortages — with low unemployment and high stress in the contact center, brands’ ability to consistently deliver high-quality customer experiences is under pressure,” Mirksic said in an email. “This new environment has created a lot of interest in deploying high-quality voice assistants that speak like humans.”
“Speak like humans” might be an overstatement – PolyAI’s voice assistant technology is somewhat limited in what it can say, to the extent that demos on the website make repetitive noises after a while. But the company says its real advantage lies in its approach to language understanding. PolyAI uses AI systems to predict what to say during a conversation with a customer, which apparently allows it to work with the “segmentation” that competing platforms require.
We can quickly deploy voice assistants without using a client’s sensitive consumer data to train the models, which means we have no risk. [systems] You are never abusing the end-user’s personal identification data,” asserted Mrkšić. “Our core models don’t require customer data for training – we work to GDPR standards wherever the customer is.”
An additional benefit of AI systems is to enable more open and “intelligent” conversations, Mrkšić said. Instead of relying on keywords and decision tree-type flows, PolyAI can interpret the context of the query (to some extent) and attempt to provide an appropriate response.
For example, if a customer calls a pizza joint and asks to update the status of their delivery order, PolyAI can recognize the phone number and proactively retrieve the relevant information. Or if a hotel guest calls to book a standard room but changes their mind and decides on a room, the platform can suggest rooms that fit the new criteria.
“Where other vendors limit how callers speak, limiting them to specific keywords and tree-style conversation flows, PolyAI has developed a comprehensive technology stack so we can understand callers regardless of what they’re saying or what they’re saying. A conversation that lasts as long as it takes to resolve customer questions,” said Mirksic.
Beyond that, PolyAI can complete many tasks that a customer service representative can, including capturing payment information as well as names, addresses, and account numbers (for health insurance, for example). Performance may vary depending on a person’s accent – while PolyAI says its technology is accent-agnostic, it’s well-documented that even the best voice recognition systems struggle with certain accents. Regardless, Mirkic says that for most customers, the PolyAI platform responds reliably until they feel comfortable — speaking like a human.
Image Credits: PolyAI
“We’ve found that people try to answer robotically because they’re used to the limited capabilities of voice technology. But, after two or three successful turns, where the customer is well understood and the voice assistant provides a useful answer, callers naturally become more open to speaking.”
The question is whether it is convincing. Most of them Get customers to change their minds about automated call handling. In the end, it may not matter—with the massive layoff push and growing call center industry, companies may embrace greater automation whether customers like it or not.
According to a 2020 Canam Research survey, 78% of US-based contact centers plan to deploy AI in some form within the next three years, to reduce average handling times and reduce customer wait times. The main point for most, unsurprisingly, in 57% of centers with AI, their goal was to reduce costs.
Incidentally, PolyAI says it has dozens of corporate and government customers whose callers interact with the platform’s voice assistants millions of times a month.
“The labor shortage caused by the pandemic has led many companies to expand their plans for voice automation. Taking the hospitality industry as an example, many workers left the industry during the pandemic,” Mikšić said. “After this, it was difficult to hire workers for hotels and restaurants. PolyAI is helping these businesses pick up calls so that their staff can assist guests on the spot.
Misterskitch complained about plans to expand the workforce – perhaps reluctant to make too many promises in an uncertain economic climate. PolyAI currently employs 98 people split between the US and the UK.
