Q&A with Verizon’s Vivek Gurumurthy: The rise and rise of algorithmic transformation
How and why is Verizon using AI to enhance customer experiences? And what does it mean to move from digital transformation to ‘algorithmic’ transformation?
23 Sep 2019
Q&A with Verizon’s Vivek Gurumurthy: The rise and rise of algorithmic transformation
At TM Forum Digital Transformation North America in Dallas today, Vivek Gurumurthy, Senior Vice President and Chief Information Officer, Verizon Consumer Group, explained during his keynote presentation that artificial intelligence (AI) is not an option for telcos – it’s mandatory. Read some of his insights in this interview about how companies are evolving from digital transformation to algorithmic transformation.
Consumers have come to expect that businesses are prepared to serve them as and when it’s needed. Beyond being reactive to their questions and needs, they expect companies to anticipate their needs and to have already attended to those needs before they know they even have them.
“As we look at how customer engagement ecosystems are evolving, we see that it’s all about having the right answer for the customer, and we have to provide it to them on their terms, where and when they want to interact with us,” Gurumurthy says. Here he explains how and why Verizon is using AI to enhance the customer experience, and what it means to move from digital transformation to ‘algorithmic’ transformation.
What must companies consider when embedding AI into their corporate strategies?
So, we’ve been using AI in multiple areas where we think we can build some expertise, but AI has a greater application to customer management than most other areas because of the amount of data available. Customer data will make AI smarter, and we use this data to ensure that we can build the right products, the right experiences, and enable the right service models for our customers.
Looking forward, our traditional OSS/BSS systems [operational and business support systems] and digital ecosystems (with respect to interacting with customers) will now be transforming into an algorithmic model. In fact, I’d say we are now moving from a digital age to an algorithmic age and how we use AI is going to directly impact the customer experience.
In terms of our business, today we are known for our network. Tomorrow, we want to be known for our experiences too, and AI will help us define our path to get there.
How does algorithmic transformation compare to general digital transformation?
Digital transformation has multiple connotations, and our own initial digital transformation involved a move from paper to digital. Then customers were doing e-commerce with us through desktop browsers, followed by interaction through mobile phones. But, in all these phases of the digitization, we saw digital as an augmentation of our existing channels of customer interaction. It was supporting the existing business models, and in many cases, was not replacing or substituting the existing rules and systems and processes. It was automating the processes that we had today, and allowing them to be exposed to the customers via certain devices.
However, now we’re looking to interact with customers on any device, on their terms. When appropriate, we use the customer data that we gather from the various touch points to write models that can then parse all that data, accurately interpret the historical patterns, help create new use cases on how the customer would want to interact with us, and open up a lot of new customer engagement opportunities for us. By using the necessary data and AI algorithms, we can potentially reshape the traditional channels of engagement, reshape traditional business models and reshape customer experiences.
Digitization was a prerequisite; algorithmic is the next phase of technology evolution to help businesses go to the next level.
And technology and AI can enable all of this?
Absolutely! All of this is possible only because of where technology is today. AI is not new; it has been in existence for the last 40 to 50 years. What is making the advancement of AI practical now is the progression of compute, software and data technologies, along with massive storage and the ability to write programs and algorithms that can compute all of this. And for us to be able to use AI effectively, this all had to come together. The timing is just right for AI to become a mainstream technology and an essential ingredient of how we interact with our customers going forward.
How important is it to an effective digital ecosystem to be able to work well with AI partners?
The way I see it is that innovation is everywhere. From analyzing current innovation patterns, every startup I see coming up in the ecosystem has a niche use case. To achieve speed to market without having to reinvent the wheel, we must have a highly active ecosystem which taps into all these new age companies; in particular, the ones that are building AI models for specific use cases. With the advent of public cloud and extreme compute on demand, the on-prem model is giving way to a dynamic ecosystem of innovation for small startups which in turn are being adopted by companies quickly via APIs.
So, it’s a completely different, very active ecosystem and is necessary to keep up with the customer needs and demands, and the shape and the speed of innovation. The key thing is – everything doesn't need to be built by us. We should have an active way of building things, and at the same time, an active way of buying.
How specifically is Verizon using AI to interact and engage with customers?
We have systems that enable customers to interact via the phone, chatbots, voice-enabled interfaces like Alexa and Google, and so on. As such systems evolve beyond the traditional ordering systems, we’ve had to build a suitable platform that seamlessly hands off the customer from one point of engagement to the other. To do this, we moved from a channel-centric architecture to journey-centric architectures where customers can seamlessly start anywhere, continue anywhere, and finish anywhere. Such omni experiences required us to introduce AI into the mix, and we now need to bring in a completely new dimension of an insights platforms that will power the customer experiences.
AI is going to help us drive the end-to-end customer journeys, define the omni experience, and build the next best actions, offers, and communications. AI is going to help us build a unique service experience for our customers and proactively predict the customers’ impending needs, be it a service interruption, a prediction on the next billing question, or a prediction on any other form of interaction they have with us for tech support, troubleshooting, additional products and services, etc.
AI is going to be fundamental to the extent that it will allow us to predict the availability of 5G to our customers when they try to qualify for service, and will proactively detect the next process to automate. AI is going to streamline our network and our systems performance and there will be AI operations and AI quality control. There are not many areas where AI can’t play a role. The OSS/BSS serving our customers will be AI powered, along with the security, fraud detection, quality control, operations, availability and infrastructure. It's going to be a fabric that stretches across the board for our services.
As a company, what benefits do you expect to see from your use of AI?
The biggest thing is that customers will love us for anticipating their next need in their relationship with us. That’s the ambition. We will serve the customer before the customer asks to be served. Of course, everything eventually boils down to improvements in churn, growth, new products, new business models, but the big thing is for us to redefine the customer experiences, and improvements in the rest will inevitably follow.
What’s the biggest challenge you’re facing in terms of AI?
The big question mark is – how do we scale AI? Building AI in a controlled environment is much simpler than building and injecting it into a brownfield ecosystem like ours. The challenge and journey here is to reinvent the traditional architectures and create new blueprints for scaling AI architecture.
Furthermore, this journey requires not only reinventing the architecture, but also retraining our employees, changing the way they work, and even changing how our teams are set up. And, we need to build an ecosystem for our teams in such a way that innovation becomes commonplace. How we take our employees along this journey will be a crucial part of how successful we will be.
As we move forward, something that remains at the back of our minds is that technology today automates the business processes in many industries, but technology tomorrow will enable new business models too, and that is a big shift that we need to undergo.
Consumers have come to expect that businesses are prepared to serve them as and when it’s needed. Beyond being reactive to their questions and needs, they expect companies to anticipate their needs and to have already attended to those needs before they know they even have them.
“As we look at how customer engagement ecosystems are evolving, we see that it’s all about having the right answer for the customer, and we have to provide it to them on their terms, where and when they want to interact with us,” Gurumurthy says. Here he explains how and why Verizon is using AI to enhance the customer experience, and what it means to move from digital transformation to ‘algorithmic’ transformation.
What must companies consider when embedding AI into their corporate strategies?
So, we’ve been using AI in multiple areas where we think we can build some expertise, but AI has a greater application to customer management than most other areas because of the amount of data available. Customer data will make AI smarter, and we use this data to ensure that we can build the right products, the right experiences, and enable the right service models for our customers.
Looking forward, our traditional OSS/BSS systems [operational and business support systems] and digital ecosystems (with respect to interacting with customers) will now be transforming into an algorithmic model. In fact, I’d say we are now moving from a digital age to an algorithmic age and how we use AI is going to directly impact the customer experience.
In terms of our business, today we are known for our network. Tomorrow, we want to be known for our experiences too, and AI will help us define our path to get there.
How does algorithmic transformation compare to general digital transformation?
Digital transformation has multiple connotations, and our own initial digital transformation involved a move from paper to digital. Then customers were doing e-commerce with us through desktop browsers, followed by interaction through mobile phones. But, in all these phases of the digitization, we saw digital as an augmentation of our existing channels of customer interaction. It was supporting the existing business models, and in many cases, was not replacing or substituting the existing rules and systems and processes. It was automating the processes that we had today, and allowing them to be exposed to the customers via certain devices.
However, now we’re looking to interact with customers on any device, on their terms. When appropriate, we use the customer data that we gather from the various touch points to write models that can then parse all that data, accurately interpret the historical patterns, help create new use cases on how the customer would want to interact with us, and open up a lot of new customer engagement opportunities for us. By using the necessary data and AI algorithms, we can potentially reshape the traditional channels of engagement, reshape traditional business models and reshape customer experiences.
Digitization was a prerequisite; algorithmic is the next phase of technology evolution to help businesses go to the next level.
And technology and AI can enable all of this?
Absolutely! All of this is possible only because of where technology is today. AI is not new; it has been in existence for the last 40 to 50 years. What is making the advancement of AI practical now is the progression of compute, software and data technologies, along with massive storage and the ability to write programs and algorithms that can compute all of this. And for us to be able to use AI effectively, this all had to come together. The timing is just right for AI to become a mainstream technology and an essential ingredient of how we interact with our customers going forward.
How important is it to an effective digital ecosystem to be able to work well with AI partners?
The way I see it is that innovation is everywhere. From analyzing current innovation patterns, every startup I see coming up in the ecosystem has a niche use case. To achieve speed to market without having to reinvent the wheel, we must have a highly active ecosystem which taps into all these new age companies; in particular, the ones that are building AI models for specific use cases. With the advent of public cloud and extreme compute on demand, the on-prem model is giving way to a dynamic ecosystem of innovation for small startups which in turn are being adopted by companies quickly via APIs.
So, it’s a completely different, very active ecosystem and is necessary to keep up with the customer needs and demands, and the shape and the speed of innovation. The key thing is – everything doesn't need to be built by us. We should have an active way of building things, and at the same time, an active way of buying.
How specifically is Verizon using AI to interact and engage with customers?
We have systems that enable customers to interact via the phone, chatbots, voice-enabled interfaces like Alexa and Google, and so on. As such systems evolve beyond the traditional ordering systems, we’ve had to build a suitable platform that seamlessly hands off the customer from one point of engagement to the other. To do this, we moved from a channel-centric architecture to journey-centric architectures where customers can seamlessly start anywhere, continue anywhere, and finish anywhere. Such omni experiences required us to introduce AI into the mix, and we now need to bring in a completely new dimension of an insights platforms that will power the customer experiences.
AI is going to help us drive the end-to-end customer journeys, define the omni experience, and build the next best actions, offers, and communications. AI is going to help us build a unique service experience for our customers and proactively predict the customers’ impending needs, be it a service interruption, a prediction on the next billing question, or a prediction on any other form of interaction they have with us for tech support, troubleshooting, additional products and services, etc.
AI is going to be fundamental to the extent that it will allow us to predict the availability of 5G to our customers when they try to qualify for service, and will proactively detect the next process to automate. AI is going to streamline our network and our systems performance and there will be AI operations and AI quality control. There are not many areas where AI can’t play a role. The OSS/BSS serving our customers will be AI powered, along with the security, fraud detection, quality control, operations, availability and infrastructure. It's going to be a fabric that stretches across the board for our services.
As a company, what benefits do you expect to see from your use of AI?
The biggest thing is that customers will love us for anticipating their next need in their relationship with us. That’s the ambition. We will serve the customer before the customer asks to be served. Of course, everything eventually boils down to improvements in churn, growth, new products, new business models, but the big thing is for us to redefine the customer experiences, and improvements in the rest will inevitably follow.
What’s the biggest challenge you’re facing in terms of AI?
The big question mark is – how do we scale AI? Building AI in a controlled environment is much simpler than building and injecting it into a brownfield ecosystem like ours. The challenge and journey here is to reinvent the traditional architectures and create new blueprints for scaling AI architecture.
Furthermore, this journey requires not only reinventing the architecture, but also retraining our employees, changing the way they work, and even changing how our teams are set up. And, we need to build an ecosystem for our teams in such a way that innovation becomes commonplace. How we take our employees along this journey will be a crucial part of how successful we will be.
As we move forward, something that remains at the back of our minds is that technology today automates the business processes in many industries, but technology tomorrow will enable new business models too, and that is a big shift that we need to undergo.
