“Conversational AI is the need of the hour. It has helped several enterprises break the economic thresholds and took off some burden from the human agents, “according to Animesh Samuel, co-founder and CEO of Pune-based AI platform E42.
Describing new-age technology as a great enabler, Samuel stressed that AI can go beyond seamless conversations and possess the ability to engage, execute, and generate actionable insights for enterprises.
Mr Samuel tells Ozonetel.com about the role of conversational AI in improving customer experience and what the future of artificial intelligence looks like.
Please tell us about the role of conversational AI in making customer services smarter. What, according to you, are the biggest advantages of deploying conversational AI for customer services?
When you buy a product or service, you need support for that. In an ideal world, we’re trying to get to a place where your product should be self-explanatory. The product should have all the information from its utility to self-service to upgradation. There should be very rare exceptions when customers have to really get back to the company. However, the fact remains even if we try to make our products more intuitive, the demand for customer care services is always going to be there, it only keeps getting higher due to new user adoption and Conversational AI plays a crucial role in improving customer experience. In today’s scenario, an efficient conversational AI and excellent customer engagement go hand in hand.
Let’s take telecom for example. Due to intense competition, the prices for telecom services in India are low, so the profit margins are very less. And now, if I have to dedicate customer service executives to manage all of that, the math will just not add up because you will be making Rs 299 or Rs 250 per month from a customer. That will be the cost of an agent talking to the customer. It will be way more expensive if we add the price of setting up Cisco Systems, routers and their maintenance cost. So, the economics will never work out. That’s where you need AI to come in and help. In the last five years, we have seen the use of chatbots picking up. These are rule-based chat engines that learn the rules like a call centre agent does. Now, these chatbots are becoming relatively intelligent. Many conversational AI players use a little bit of AI in one part called intent classification which helps them understand what the customer is trying to say. It’s a win-win for both sides. On one hand, customers are not left out in the lurch and enterprises get to improve their services.
AI is no longer a fringe technology. We have seen a lot of traction in the last couple of years, and the global conversational AI market has grown rapidly. How is the market for Conversational AI/AI chatbots growing in India?
Conversational AI is the need of the hour. This space has seen an upward trend over the last few years and it’s here to stay. It’s the technology which makes conversational AI irresistible. In the 90s, we used to listen to music through tape recorders and then came the CDs which could carry more than 250 songs on a single disc. It allowed us to switch to any song of our choice within seconds. Similarly, conversational AI is an upgraded tech which will see more adoption unless a superior technology comes in.
During the pandemic, many companies were under a lot of pressure from customers, due to covid-induced lockdown. This led to massive digitization. Do tell us about the impact of the pandemic and how it propelled the growth of conversational AI space.
During the pandemic, people were not able to travel to work. It created a massive demand for support on both sides. We called it (pandemic) the new ‘chief digital officer’ because it enabled digitization across enterprises. In AI, ours is a no-code platform on which one can build AI workers that not only converse but also take action, based on different requirements. We have created AI workers to engage with customers exclusively and that’s where the conversational AI part comes in. We have seen the adoption of a large number of AI workers not just in customer care but also across other segments. We have built AI gatekeepers, AI analysts and even AI HR executives among others. We all have experienced how it is to repeat the same problem to a customer care executive and then go to a supervisor, and so on. All of those things never happen with AI workers, they don’t forget anything. In fact, now you reach out with one complaint and a year down, it can actually take on that conversation from a different channel. So, if you said your first complaint was on WhatsApp, one year later on Facebook Messenger, the new technology can respond on both platforms, it will be a continuous seamless conversation. So, the advantage of an AI worker over a conversational AI is that it can execute tasks, which also becomes a learning for other departments.
Customer interactions a decade from now will probably look and sound quite different. The new algorithm might need to understand new words or anticipate what consumers want? What, according to you, are the major barriers to the growth of conversational AI and how can we overcome them?
Let’s be very clear if it is AI, there will not be many algorithms in it. You feed into a system and AI learns on itself. If there is coding involved, then it’s not AI. It also allows us to tweak the data, for example, for a set of data that we train our models on, we remove the identity because we don’t want any identity bias in the AI. So, you massage the data and send it in and the rest will be done by AI. There are different methods that can be employed. AI comprises neural networks that act like a brain. The way we react is not because someone told us to react this way, it is how our brain is wired. Similarly, AI is just replicating that. So, there is no code, and there is no algorithm. So, if the language changes, if the way the new generation interacts changes, all that can be easily adapted due to AI, in fact, it’s much more challenging for a human, sitting in India to understand the changes in language in the US. For AI, it’s just a few million transactions, and it will pick that up itself. So, AI presents us with a great opportunity that goes beyond a particular industry.
An Accenture Strategy study found that 48% of people want special treatment for being loyal customers. How conversational AI can help businesses in customer retention and acquisition. retain old customers and find new ones.
AI can be our best bet for customer retention. It can serve as an efficient call center executive through data. For example, if you find a fault in your product and want it replaced. An AI-enabled agent can smoothly execute it without the need for human intervention. It doesn’t have to say let me raise the ticket and make the customer wait for another day. Having said that if the person happens to be an old customer, we should walk the extra mile to assure him that it would not happen again and how valuable he is to the company. We can always have a small percentage of humans in the loop also, when the AI is not able to give a good answer, we can bring them to complement those systems.
Which segments are the fastest movers in the Conversational AI space, and how has it helped them in improving the overall customer experience?
We have seen a lot of adoption in the BFSI and telecom sectors because the margins are very small in these segments and the cost-benefit is high when you deploy conversational AI. But conversational AI was been there for the last five and six years, now, we are talking about AI workers or digital twins that can, in fact, make a better version of a human to do the job. So, it is no more about conversing in a pre-decided manner, but it is also about taking action and reporting that action, which makes all the departments smarter.
Putting in place a system to manage customer conversations is just one part, equally important is extracting value from the data provided. How brands can leverage conversational analytics to derive actionable business insights from customer data.
You can leverage conversational AI to get business insights in real-time when the conversations are happening, and when actions are being taken. So, for example, customers are searching for the best interest rate for a fixed deposit and have found offers in certain regions to be lucrative. That knowledge in real-time is an insight, which could have been difficult for humans because it’s AI which can find and give that insight to the business. Using these insights, the marketing department can push those kinds of messaging in those regions and get a much better outcome. So, if it’s an intelligent system, it has to go beyond conventional. When it comes to enterprise systems, it has already witnessed the evolution of chatbots, conversational AI and Robotic Process Automation. But now what we are seeing is cognitive process automation which is what our organization does.
What suggestions would you give to a company that’s working to deploy conversational AI for the first time?
You can deploy conversational but then you should also consider deploying more than that. So conversational AI has its benefits but you have to make sure that IT systems are intelligent, not just at intent level but it should be able to use data from other departments as well. As a first-time user, you have to ensure that the intelligence is captured through customer interactions and it should only be used on your cloud instance. Another important part that needs to be factored in is its ability to capture data in real-time and convert it into knowledge graphs that can be used by other departments. Today AI can go beyond seamless conversations and has the ability to engage, execute and generate insights.
Machines and AI were actively deployed to suffice for the job losses due to covid-induced lockdown. This perception was created that emerging technology can replace humans but now a lot has been said and written to debunk this perception. Just wanted to know your perspective on the issue?
The answer lies in the aftermath of the Industrial Revolution. When machines were invented, a perception was created that the machines would take away all our physical jobs but many more jobs were created in making those machines. Similarly, when the computer gained prominence, people said computers will take all our clerical computational jobs. They took away some jobs but many more were created in making computers, software and hardware. The same question is being asked again. There will be ample opportunities in building AI. We were born to think that we had to take a nine to five job but that concept emanated after the industrial revolution. We are 3 lakh-year-old species and AI is still in its infancy, how it impacts work systems, in the long run, is yet to be seen. AI will take care of most of the industrial and enterprise work but yes it will be complementing humans.