Google is bringing its human-like conversational AI to the contact centre

Google Cloud has announced a new enterprise contact centre solution which uses AI to mimic humans to answer simple queries and automatically route callers to the right agent.

Scott Carey Jul 26th 2018
google-next.jpg

Google Cloud has announced a new enterprise contact centre solution which uses AI to mimic humans to answer simple queries and automatically route callers to the right agent.

The solution - called Contact Centre AI - was announced on stage during Google Cloud Next in San Francisco this week and builds on Google’s Dialogflow Enterprise Edition, its development suite for building conversational agents that was announced back in November. This has been enhanced using DeepMind’s WaveNet, which can then be integrated into telephony networks using the Dialogflow Phone Gateway. 

WaveNet is “a deep generative model of raw audio waveforms” developed by the Google's UK subsidiary DeepMind and announced in 2016. It promises the ability to “generate speech which mimics any human voice and which sounds more natural than the best existing Text-to-Speech systems, reducing the gap with human performance by over 50 percent.”

In a blog post, Fei-Fei Li, chief scientist at Google AI wrote: “When we studied the challenges faced by real contact centres every day, we found that customers often have simple transactional or informational requests.

"For contact centre employees, this can mean repetitive work, increased pressure on caller turnover, and less time to solve complex problems. For the callers themselves, it can mean frustrating menus, longer hold times and a diminished experience overall. This got us thinking about how we could use AI to dramatically improve the experiences of both customers and the contact centre representatives that help them.”

In practice Contact Centre AI works by replacing the old phone tree system with a virtual agent that answers questions before handing over to a human agent. Contact Centre AI also helps support agents with contextual information so that they can avoid asking for the same information again.

The enterprise solution is “compliant with our data privacy and governance policies, and it does it all without infrastructure, on a platform that scales as much or as little as you need,” Li wrote.

eBay case study

An early beta customer of the solution is eBay, which has been working on a prototype solution with Google Cloud and its contact centre technology provider Genesys to use a natural language bot to eliminate the dreaded phone tree.

Speaking on stage during Next Dan Leiva, VP of customer service at eBay, said that the aim of the pilot is to solve three problems: “First is when a caller has to navigate the infamous phone tree and answer an endless stream of questions just to get the system to recognise the nature of the call. Most of the time you just want to talk to a human being.

"The second problem is keeping track of information, half the time we finally reach a human agent they ask us to repeat information we just gave the phone tree. Lastly, agents often have limited access to the information they need on that caller.”

Merijn te Booij, CMO at Genesys added: “We like to greet our callers with natural language, we don’t see any reason for this paradigm of answering yes/no questions." Genesys already provides a predictive call routing system, so combining this with Google’s AI solution smooths out this process even further.

As an aside Twilio’s cloud contact centre solution Flex has also announced that it will support Google’s new AI solution.

Other initial partners include Appian, Chatbase, Cisco, Five9, Genesys, Mitel, Quantiphi, RingCentral, UiPath, Upwire, and Vonage to develop best practices, “such as disclosing when customers are talking to a bot, and education around issues such as unconscious bias and the future of work” Li wrote.

Naturally, Google is keen to present this as a human-augmenting, not replacing, technology.

Li writes: “With intelligent tools that offload repetition and support human strengths like creativity and problem solving, we're demonstrating that AI’s greatest potential lies in complementing and enhancing human skills, not replacing them.”

Duplex similarities

The technology bears a striking resemblance to Google Duplex, an AI application that can conduct human-like conversations over the phone which caused some controversy when CEO Sundar Pichai demoed it earlier this year at the I/O developer conference. The criticism centred on the AI not clearly identifying itself as a bot.

In a demo of the Contact Centre technology along with eBay on stage during Next, the agent immediately introduces itself as a virtual agent, showing that Google might well have taken this feedback on board.

Responding to Computerworld UK’s question about this similarity, a Google spokesperson responded: “Contact Centre AI is not based on Duplex. Contact Centre and Duplex share some underlying components, but have distinct technology stacks and aims overall.

“Contact Centre is focused on the enterprise and Duplex is focused on a limited set of consumer experiences. Contact Center AI is compliant with Google Cloud’s data privacy and governance policies, making it ready for enterprise use immediately.”

Contact Centre is available in alpha and is open to sign ups immediately.