Home > Columns > Executive Interviews

Vonage Executive Interview

Savinay Berry, EVP of Product and Engineering, Vonage

Click image below to view the ebook

ebook image

Sheri Greenhaus, Managing Partner, CrmXchange, met with Savinay Berry, EVP of Product & Engineering at Vonage at Enterprise Connect to discuss the advantages of Vonage’s communications platform and their new AI acceleration suite.

The CrmXchange audience is dedicated to helping their customers. Customers are willing to use bots as long as they can quickly and efficiently solve their issues. However, if the chat or bot is unable to solve the problem, they want a quick and contextual escalation to a human agent.

I slightly disagree with the term "escalation" and prefer to think of it as a two-way handshake between the bot and the human agent. This is to ensure that the transfer between the two is seamless and so the customer cannot tell the difference.

As a communications platform provider, we have the advantage of having all channels enabled as part of the platform. This makes it easy for them to take any channel of choice for the customer, provide support for the application, and seamlessly transfer the dialogue in context. This is in contrast to pure contact center providers who only provide specific contact center applications. 

Many businesses lack a seamless transfer from chat to agent.

Yes, the reason for this issue is the lack of shared context between the customer's chat and call interactions. When a customer starts with a chat but doesn't find a solution, they may resort to calling the company. However, from the company's side, there is no connection between the two interactions, resulting in a terrible customer experience. This problem is challenging because it requires data normalization to identify the original person and semantically link the information in the backend to a single data source. Many people call this data storage the CDP. 

We are creating a behavioral and intelligence engine using AI components, such as new language models. This engine will be able to identify the customer at scale for billions of people across hundreds and thousands of companies, despite variations in data storage maturity and how marketing tools connect with customer service tools.  

Should there be one single source of truth for the data?

This is the idea behind the CDP, which provides a 360-degree view of the customer. Our goal is to provide a productized approach to solving this problem for 80% of customers. The last 20% can be customized based on the customer's environment and the tools they use across marketing and sales.

Who do you target?

Our legacy comes from New Voice Media, which was deeply integrated with Salesforce. The sweet spot is the Salesforce install base for contact center and customer service with a range from 500 to 5,000 agents. There is great importance of integrating Salesforce Service Cloud voice with UC. 

We are starting to see not just the contact center, but rather the entire employee base answering questions on behalf of customers seamlessly. We saw a large majority of the base transferring call queues to an individual with the right skill set, which becomes important. 

There are two problems that need to be solved. The first problem is how to enable the tools that connect the employee base and the support center, making it a single entity. The second problem is how to connect the contact center or customer experience data with the marketing data.

What percentage of companies are currently solving these two problems?

A very small percentage of companies are doing this, and that's where we’re investing to solve these problems. We recently announced AI Acceleration Suite, a portfolio of AI-capable, low-code/no-code programmable components that speed and simplify the development of applications. AI Acceleration Suite allows customers to take local applications and build capabilities to integrate with their existing data, unlocking and normalizing it.

Who would be best to operate your tool?

It may be an operations manager who understands the business and desired metrics, but an IT background is not necessary. Our tool enables engagement with customers in multiple channels and reduces call handle time. The call may never even reach the agent as the issue was solved without human intervention.  AI Studio, for example, which is included in the AI Acceleration Suite,  is a drag and drop tool that designs the workflow. All one needs to understand is the business processes.

Why don’t more companies use this type of tool?

Many decisions are being made myopically with thoughts only for the current year rather than considering how to transform their entire in-house experience. People start to wake up only when there is a significant drop in CSAT or another metric.

We want to change that mindset of the customer-business communication; one that makes it feel like customers are talking to a friend. The goal is to create a conversational engagement where the customer is, whether over WhatsApp or other channels. 

People reach out to a brand for two reasons – to buy something or to get help.  It should be an easy ‘conversation.’ This emotion that they are trying to achieve should be one where the customer feels like they are talking to a friend. That's what’s driving a lot of the work we are doing in our contact center. We think of it more as conversational engagement and we're finally starting to see what that human like experience might look like. 

To some extent, talking to a chatbot still feels like talking to a robot and getting a simple yes or no answer. Our vision is to turn notifications into conversations, on whatever platform the customer prefers, enabled by the world’s largest engagement platform. It’s a simple belief that as human beings, we want to have conversations and engage with others.

When it comes to service, most customers don’t know or care so much if it’s a bot, they just want you to fix the problem.

That’s it. They really don’t care. Having that solution lies with people like us. They don’t care if there’s a handoff happening from a bot to a human or human to bot – they just want someone to solve their problem. What is happening behind the scenes does not matter to them.

Do you find people are concerned that AI will take agents' jobs?

There’s a new category of work that's going to start emerging called prompt engineering, which is basically a way to devise the right prompt to get the right answer from these large language models. And it's almost written in English prose, which you need to get just right.

You can't really write the right prompt if you don't know the business and you don't understand what the issues might be. If you really want to go deeper and customize the company’s experience, you really need to know the business.

In my view, contact center agents, especially the ones that have been in the business for a while, help engineering work to train the model. They don't need to understand coding. They just need to understand the business and how to train the bots. Agents will start to work with the hardest and most complex issues. That's where escalation might come in as agents receive the more challenging calls. And third is, agents will provide white glove treatment for select high-value customers.

I think those three agent paths will start to emerge. With our low code/no code solutions, when you have the agent well-versed in the business, they can be instrumental in designing the workflow.

What is the overriding problem that you are solving?

Getting to where the billions of people who need customer service, whether it's from the flower shop down the road or big insurance or financial companies, literally feel like they are talking to a friend. Ultimately, products that are out there, they're less about the brain, it's more about the emotion that you feel when you're connecting. 

As we have discussed, the issues are: How do you chat and get your answer very quickly? How do you get the right answer based on the right channel that the customer prefers? How do you know what the customer prefers? That's where the intelligence part comes in. The way to solve that is going to be the customer data platform and all the intelligence that I mentioned earlier. But it's ultimately about how billions of people across hundreds and thousands of companies can interact to feel like they're talking to a friend.  We believe conversations are going to become the next digital storefront and it does not matter the size of the company.

The same exact tools that some of the world's largest companies use should also be available to a mom-and-pop shop down the road where they don't have developers. That's the vision we have. And that's the beauty of having an end-to-end communications platform.