Google Brings AI to the Contact Center

By Blair Pleasant
30 Jul 2018

At the Google Cloud NEXT conference in San Francisco, Google made many, many AI announcements, including new AI tools to help contact centers be more efficient and effective. One of the major announcements is the Google Cloud Contact Center AI solution, a “simple, secure, and flexible solution that allows enterprises with limited machine learning expertise to deploy AI in their contact centers.”

Google Cloud Contact Center AI includes the Enterprise Edition of Dialogflow, a development suite used by developers and vendors to build conversational interfaces and chatbots for websites, mobile applications, messaging platforms, and IoT devices to improve the customer experience. Google added some enhancements to Dialogflow Enterprise Edition, which can now build AI-powered virtual agents for contact centers, and also includes a Phone Gateway for taking calls, and Knowledge Connectors for understanding unstructured data like FAQs and Sentiment Analysis.

Powered by Google machine learning (ML) and natural language processing (NLP), companies can build conversational interfaces for customer service use cases where virtual assistants or agents can schedule appointments, answer common questions, or perform basic tasks. The system continually learns and improves over time.

The Contact Center AI portfolio has several components that can be leveraged by Google’s contact center partners and customers:

  • Virtual Agent: a conversational chatbot that can detect customer sentiment and be used for interacting with customers. It works with a company’s knowledge base and back-end processes.
  • Agent Assist: for live agent escalation. When calls are sent from the virtual agent to a live agent, Agent Assist suggests answers and provides information to agents based on previous interactions. This makes the agent more knowledgeable in their role to better resolve the customer’s issue or answer their questions.
  • Analytics: Using NLP, the AI Analytics can be used to look at large data sets to: identify trends, identify why customers are contacting a company, coach agents, and improve agent performance.

Here’s a typical scenario of how these components work together: The Virtual Agent can automatically answer and respond to customer calls and complete basic tasks. If the caller needs to speak with a live agent, Virtual Agent passes the customer to a live agent, where Agent Assist provides the agent with relevant information to help better respond to the customer. Using natural language processing, the AI solution can be used to analyze conversations in real time and find relevant documents to present to agents while they’re interacting with customers to better serve customers’ needs. Using Dialogflow’s Knowledge Connectors, the most relevant information from a company's knowledge base are presented to the agent, helping them resolve the customer’s inquiry.

Partners are Key

Google is partnering with several contact center vendors that are leveraging the Google Cloud Contact Center AI solution, including Appian, Chatbase, Cisco, Five9, Genesys, Mitel, Quantiphi, RingCentral, Twilio, UiPath, Upwire, and Vonage.

Several of Google’s AI partners are also working with Google to develop and share best practices when it comes to using AI. According to Mitel, which is engaged with Google as a flagship partner of “Responsible AI,” “Google recently announced a set of ​AI Principles​, which set out Google’s commitment to develop technology responsibly and establish specific application areas they will not pursue.” These principles include issues such as educating businesses around issues such as unconscious bias.

Quick Contact Center Vendor Overview

Remember the children’s game where you try to find the differences between two similar pictures or drawings? Many of the contact center vendor press releases announcing the Google partnerships were very similar, focusing on the partnership rather than actual offerings that will be based on the Google technology. It’s clearly too early to tell how some of the vendors will be leveraging the Google AI technology, and I expect that over the next few months we’ll see actual product offerings. It’s interesting to note that many of these vendors will also be competing with Google’s new enterprise version of Google Voice for G Suite users, and we’ll have to see what happens to some of these partnerships as Google encroaches more on the enterprise voice territory.

To try to differentiate the various vendors’ approaches, I asked several of Google’s contact center partners how they’ll be using the Google Cloud Contact Center AI solution, and what they’re doing that’s different from others. Here are some of the responses.

Cisco: Using Google Cloud's new Contact Center AI solution, Cisco announced that it is adding Google AI to its Cisco Customer Journey Solutions, combining Google's AI capabilities with Cisco's large global reach. The Cisco team has been working with Google for some time, and Cisco was an early adopter with API.AI which subsequently became Dialogflow. The new Contact Center AI project began when Google approached Cisco almost a year back to partner on the definition of the agent assistant service. The Cisco relationship with Google spans multiple billion dollar product categories and therefore is strategic to both companies.

From a contact center perspective, Cisco notes that a key difference between Cisco and its competition in this area is scale. Cisco has developed its integration in such a way that the company will be releasing it to all of Cisco’s platforms; cloud and premises-based alike. This brings [cloud] AI technology to all of Cisco’s installed base of premises-based customers in addition to its cloud contact center offer.

In a new blog, Tod Famous, Senior Director, Product Management, Cisco Customer Contact Business Unit, noted, “We are proud to collaborate with Google and its new Contact Center AI service. With this integration, agents don’t have to search through many systems and knowledge bases to find a response to a customer request. The “intelligent agent assist” feature automates the process, increasing agent effectiveness and efficiency. It ensures that your agents can always give customers the best possible response to their queries.” He adds, “Existing customers using our premises-based platforms do not need to wait 12 months for the next release. Cisco is building cloud-based services that overlay on our premises-based contact center platforms. We are bringing cloud-speed innovation to our customer base.”

Five9: According to a Five9 press release, key features of Five9’s integration with Google Contact Center AI include:

  • A natural and engaging conversational virtual agent designed to provide consumers a way to quickly resolve more common issues without need to speak to a live agent
  • Easily escalate a call to an appropriately skilled live agent with all relevant data from the self-service interaction
  • Live agent augmentation, designed to guide and assist the live agent to deliver more prescriptive, accurate and timely responses. This is accomplished using Google Cloud Agent Assist technology.

Five9 notes that the Google partnership further extends the power of Five9 Genius, which was designed to take advantage of innovations in AI from companies such as Salesforce, IBM, and Amazon. It allows Five9 to use its contact center knowledge to solve customer issues by leveraging a wide range of market technologies. The company states, “The Google partnership is another example of how Five9 can leverage the investment in AI to help solve contact center problems.”

The company adds, “The key for Five9 is that we’re a cloud-based vendor that has access to large numbers of customer recordings. We have a very rich repository of data which is a requirement for AI training.”

Genesys: It was nice to see Genesys CMO Merijn te Booij on stage at the Google Next event presenting a demo of how Genesys is using Google Cloud Contact Center AI. By integrating Kate, the Genesys AI platform that powers the company's natural language understanding, orchestration, predictive routing, automation, and more, with the natural language understanding capabilities of the Google Cloud Contact Center AI, the AI system can detect what consumers are saying and understand their intent.

In a blog about the partnership, Genesys Chief Product Officer Peter Graf explained, “With Genesys Predictive Routing technology and the newly launched Google Cloud’s Contact Center AI, live agents and AI can seamlessly work together to give consumers a better experience that uses automation when it makes sense and the human touch when it matters most.”

In my discussion with Genesys, the company noted that it started its AI Journey before other vendors and has been in AI platform development with Kate since 2017. “Google recognized our differentiation with Predictive Routing. Genesys was the first partner to show a real-life demo with a shared customer, eBay. Our open AI approach has been fundamental to the launch of Kate two years ago, we are continuing to bring the best to our clients in an open and collaborative way.”

The company added that bots need context and data to do work, including product knowledge, policies knowledge, customer history, activity on other channels, order history, etc. Genesys customers already have all of that connected to Genesys, and Kate can provide that data to Google (and other AI platforms).

Kate also acts as a bridge between the Google AI and all of the other channels a customer may have. This enables sharing of information across channels, so that a bot interaction on one channel can be quickly shared to other channels.

Mitel: Mitel has been working closely with Google to develop the next evolution of the contact center experience from all perspectives, including the end-customer, contact center agent, and business development. Going beyond “simply consuming Google’s APIs,” Mitel has been working on co-development with Google and providing feedback on how to improve the offering. Mitel notes that it is dedicated to improving the productivity of live agents, integrating the leading AI and machine learning capabilities from market leading partners.

Mitel notes that the Google Contact Center AI solution is “a simple, secure and flexible way to allow businesses with limited AI expertise to implement it into their Mitel contact centers quickly and efficiently.” This is helping Mitel deliver on its CloudLink strategy (which allows businesses to leverage their existing Voice and IT assets, while adding and extending new capabilities from the Cloud) by combining functions in the CloudLink platform with Google APIs to make them deliverable across the entire Mitel contact center portfolio. Mitel notes that working with Google Contact Center AI allows for real-time insights to continually improve the customer experience, while providing real-time coaching and assistance for agents.

Mitel is using Google Cloud Contact Center AI functions for AI Analytics, Virtual Agent, and Agent Assist.

Mitel customer Acendas Travel is a proof of concept showcase customer using the Mitel MiCloud Connect Contact Center and Mitel CloudLink with Google AI technology to “Improve the customer experience by personalizing interactions, enhancing support and helping our advisors be more effective in providing consultative vacation planning.” Acendas is using the technology to help customers get to the right agent (advisor) who can help them throughout the travel booking process. Using natural speech, the customer can talk to the Virtual Agent, and the Virtual Agent can even book the vacation. If and when the call is transferred to a live agent, the agent gets a transcript of the voice conversation and can see the full interaction with the Virtual Agent. Agent Assist presents the agent with information to support the conversation and interaction, and even suggest offers and packages for the customer. Acendas found an increase in success rates, as they can help customers identify their preferences more quickly. The company also had more repeat bookings based on more personalized experience.

RingCentral: RingCentral and Google Cloud are collaborating on “applying AI in multiple areas of RingCentral Contact Center, enabling deeper customer engagement and improving the productivity of agents.”

RingCentral will be using Google Cloud Contact Center AI for:

  • Self service across voice, chat and other channels; 
  • Combining input from AI with CRM data to select the best possible agent to support a customer; and
  • Providing suggested answers and assistance to the agent while working with customers.

According to the RingCentral press release, sample use cases leveraging AI insights include:

  • Customer self-service over the web, mobile, and voice
  • Sending conversation context to live agents
  • Real-time agent assistance with article and answer suggestions so customer conversations are more meaningful
  • Cross-selling and upselling

When asked what they are doing that is different from other vendors, RingCentral noted, Google AI provides powerful insight on who is calling and what they need.  RingCentral is unique in our ability to use that information to connect with people across the entire enterprise, not just in the contact center.  For example if the customer needs to talk to billing or collections and they are not part of the contact center, we can make the connection and provide a screen pop. Or if the customer is in the middle of a complex support ticket, they can be sent to a knowledge worker who can help with the issue, instead of someone in the contact center who will need to find that knowledge worker to get help.

Vonage: Vonage notes that it has built a solution that offers the ability to connect customers with an intelligent virtual agent by integrating Nexmo’s Voice API (VAPI) and skills-based communications routing technology with CRM, Google Cloud Contact Center AI, and Dialogflow, giving the Virtual Agent the real-time data needed to provide contextually relevant answers. The solution also seamlessly and contextually routes customers to the best available agent when a human touch is needed through a connection between Nexmo’s skills-based communications router and Dialogflow. Once the customer is speaking with an agent, the solution is still working as an Agent Assist interface that provides continuous, real-time transcription of the call and dynamically updates suggested responses based on the ongoing conversation.

When asked what Vonage is doing that is different from other vendors, the company noted, “Our team was inspired by the Google contact center capabilities to create a product and experience that takes advantage of the power of Nexmo, the Vonage API Platform, giving developers a big head start to using these technologies together to drive more meaningful customer interactions."

The new integration is being made available as part of Vonage’s CX Enablement Suite, which encompasses several new Nexmo-powered contact center capabilities including programmable building blocks, chatbots, and open source skills-based communications route and real-time sentiment analysis, and can be used to customize and enhance existing contact center solutions, including Vonage CX Cloud.

Conclusion

While it will be several months before we see the fruits of these partnerships, it’s clear that AI is the future of the contact center and customer interactions. Google has made a strong play in this area and is working with some of the leading contact center vendors. Partnering with Google, these vendors plan to bring AI capabilities to contact center customers, helping to improve agent efficiency eventually resulting in faster and better service for customers.

I’m looking forward to seeing how businesses deploy AI capabilities for customer interactions, as success will not be based on the technology, but on how the technology is deployed and used. Organizations will need to understand and deploy best practices to ensure that the AI technology actually helps improve the customer experience.

Comments

There are currently no comments on this article.

You must be a registered user to make comments