Google Contact Center AI is Available – What That Means for Your Call Center


Share on LinkedIn

On November 14, 2019, it was announced that Google’s Cloud Contact Center AI is now Generally Available (GA). A lot of the partners that Google is working with (NICE InContact, Five9, Genesys, TalkDesk, etc.) have known that this was coming for a while and have been working together to develop products for their telephony platforms.

Expivia uses NICE inContact, a leader in the cloud telephony platform, and has had lots of communication with the product team about how things are going and where they see AI really making the biggest, fastest impact on our centers. Where they go with the AI telephony platform, a lot of people are going to follow. We are excited and cannot wait to get involved with this.

At Dreamforce this year, Salesforce announced that they are going to partner with Amazon Connect to become a one platform product, that will continue to work with their current partners like InContact, Five9, and Genesys, etc.  Centers will still be able to integrate with Salesforce, but they are going to have an off-the-shelf product with Amazon Connect, allowing users to switch and not use those partners if they choose.

There are lots of ways to use the AI that is coming up.  A year ago, at a Google conference, there was a demonstration of a phone call to create a hair appointment.  The customer didn’t know they were talking to a robot/AI; while the technology may be Generally Available, it may not be Generally Acceptable yet. Some customers may really dislike talking to AI instead of a human. Don’t think that will to stop this technology from advancing and becoming more popular. Should we be telling customers that they are talking to a robot? That’s up for debate right now.

We will be able to use this AI to help our agents in the call center, but we won’t replace them with it. This summer Expivia will be bringing this technology on board.  We are already working on the dialog flows with a couple of our clients. One of them has gone all-in on this, and we’ve worked hand-in-hand to develop some AI chatbots and some natural-speech IVR (Interactive Voice Response) just for them.

Natural speech IVR is not the IVR from the early 2000s that still sounded very stilted and mechanical. Now, natural-speech IVR is able to respond, without sounding like a robot, to not just the words but to the tone and sentiment, too. Everyone wants to use natural-speech IVR, but it hasn’t completely evolved yet.

Focus is being placed on using tech to upgrade the self-service models that we use for our customers.  It will be able to fully integrate into our calls.  When a call comes in, it will be listening to what is being said and it will learn from what it hears. It will also be able to seamlessly transition to an agent when needed, which helps the customer experience.

Agent Assist is the next big piece of AI that Expivia is going to get involved with because we love speech analytics. Google AI will be doing both real-time translation and transcription. As the customer is talking, the translation side of the AI will translate what is being said instantly to the center’s local language.  Even if the caller isn’t speaking English, the agent will see what is being said in English in real-time. Being able to hear the translation in real-time should be possible soon. 

This tech will also offer real-time transcription. With real-time transcription, when an agent begins talking to a customer, the transcription is already running. The algorithms in the AI are hitting those transcriptions as the call is going on, allowing us to look at sentiment during the call. 

Currently, analytics are all done after the call. The call is recorded, it goes to a server where it gets transcribed for our analytic purposes, and then it goes to a different server where the analytics that we need are performed. The set algorithms get to work pulling out the trending keywords, checking for tone, and recording metrics that you request. Then, it presents us with the overall readout with sentiment and keyword analysis. We can do a lot with this information from  QA and marketing standpoints. Looking at this information after the call is over is not nearly as helpful as doing it in real-time.

A supervisor will have their dashboard open watching all of the current calls and will see on their screen when a call’s sentiment starts to go south. In addition to this, Agent Assist will actually give the agents help with how to service, how to sell, and what to say to the customer based upon real-time sentiment scores and processes that have been set up on the back end. For example, if the AI hears a customer say “thank you,”  it will suggest ways to cross-sell or upsell them a certain product.

In making the customer experience better, this technology will allow the customers to help us turn our cost center into a profit center. It will help our agents really drill down, sell more, convert more, and have an overall higher conversion number, leading to happier customers, growing programs, and bringing more business and money to our call centers.

There are always detractors to new things. Some are saying we are going to get rid of all customer service agents worldwide and replace them with virtual agent robots. They believe we are going to use only the AI, and it will eventually take over everywhere.  That’s not going to happen.  Early AI adopters are getting a laugh at this notion that technology and virtual agents will replace actual people.

In the near future, there will definitely be humans working in customer service call centers. The technology is just not there yet for them to be deemed obsolete. This might change with time, but we don’t plan to.  Expivia plans to keep using the really cool technology, keeping everything up to date and fresh. There will always be a place in our centers for humans. One day, our clients will start to advertise not only that they use American customer service agents, but that they are real humans, too!

This is not to say that AI isn’t going to be used on most calls. We are going to position ourselves to utilize this technology as much as we possibly can.  Using it will help the customers have a better customer experience. The next competitive battlefield is customer service, and AI will be a huge part of it.

If your organization doesn’t have the ability or desire to learn to use the AI that is coming, it may be time to look at out-sourcing. Those people will be able to work on the AI for you so you do not get left completely behind. There will be many companies able to do this. 

Do your research. Understand that you don’t have to spend lots of money on AI. There are partners out there to help you.  Right now, the early adopters are jumping on board.  Expivia has never done anything like this. We are programming all day, and we cannot wait to see it all play out. Contact Center AI is not a fad; it is here to stay. Is your center ready?

Want more call center operations content? Head over to our weekly call center operations podcast “Advice from a Call Center Geek!” at

Advice from a Call Center Geek is a weekly podcast with a focus on all things call center and contact center. We discuss topics such as call center operations, hiring, culture, technology, and training and have fun doing it! #callcenter #contactcenter #CX #custserv #callcentergeek

Thomas Laird
Founder and CEO of award-winning Expivia Interaction Marketing Group. Expivia is a USA BPO omnichannel contact center located in Pennsylvania. I have 25 years of experience in all facets of contact center operations. I have the honor of being a member of the NICE inContact ICVC Board. The iCVC is a select group of inContact customers selected to join as trusted advisors to help InContact validate ideas for new products and features and plans for future innovations. I am also the author of "Advice from a Call Center Geek" and host of the podcast of the same name.


Please use comments to add value to the discussion. Maximum one link to an educational blog post or article. We will NOT PUBLISH brief comments like "good post," comments that mainly promote links, or comments with links to companies, products, or services.

Please enter your comment!
Please enter your name here