AI Leadership Insights: Conversational AI

Amanda Razani: Hello and welcome to the AI Leadership Insights series. I’m Amanda Razani, and I’m very excited to be here today with Stefanos Loukakos. How are you doing today?

Stefanos Loukakos: I’m great, thank you. Very excited to be on your podcast.

Amanda Razani: Now, Stephanos is the CEO and Co-founder of Connectly. And can you tell me a little bit about Connectly and what services you provide?

Stefanos Loukakos: Absolutely. Connectly is a three-year-old startup, and we help businesses and especially e-commerce and retailers to communicate with their customers on any messaging platform. And we do that with the use of AI. So we fully automate the way e-commerce and retailers are selling their products to their customers.

Amanda Razani: Wonderful. Well, that is a great segue into our topic of conversation today, which is the online customer experience, especially in regard to retail. So when it comes to the online experience, what are some of the main problems that business leaders face when it comes to the customer?

Stefanos Loukakos: Great. So this is a great question and I’ll tell you a little bit of how we are thinking about this. And this might not be the best answer or the correct answer, but this is what we believe. So if you think about it today, the majority of people, like myself for example, do not communicate with businesses the way I communicate with my friends and family. So with friends and family, we always text each other. We use WhatsApp or SMS or iMessage and all these channels. But with businesses, we have to call them, right? Most of them. Or we have to go to a very personal website. And we believe that this personalization, the fact that businesses can have personalized conversation with their customers is a must, and this is something that helps them sell more products and become way more relevant to their customers. So what we’re doing at Connectly is we help businesses do exactly that. And not only that, we help them do it with the use of AI by fully automating it. So we help businesses have one-to-one conversation with our customers, personalized conversations, in order to be able to understand who their customers are and sell the right product to the right customer at the right time.

Amanda Razani: So that’s great. And I know that there’s a lot of technology out there, primarily let’s talk about artificial intelligence, which is the latest and greatest technology that everybody’s trying to harness. How does that play a role and how do they come about implementing it properly?

Stefanos Loukakos: Again, great question. And just to tell you a little bit about the history of Connectly, we didn’t start as an AI company. We started going all in on AI after the ChatGPT announcement like a year and a half ago, where we realized that the messaging that businesses have today is, no, no, the way businesses communicate with our customers today is mainly one to many. So they send the same message to as many people as possible. So it’s maybe it’s one ad or one SMS that goes to a lot of people. So it’s one to many. And what the ChatGPT and AI change is, guess what? You don’t have to have one too many messages that people do not reply back. Actually, people can reply back and you can have a conversation with your customers and you can fully automate this.
So what AI changed is the fact that businesses now can have personalized one-to-one conversation with our customers without the need of having humans actually making these conversations. So we feel this is a very exciting new space, but it’s not an easy space, and it’s not like you just use ChatGPT to make it happen. Businesses need to use their own data, they need to train their own data in order to work for the use case they want it for. We need to optimize the models for conversions, sales and things like that. So this makes it a very, very complex problem. We haven’t solved it 100% yet, but we believe we’re the right way.

Amanda Razani: So when it comes to this big project, it’s a big project all the different steps, what are some of the big roadblocks that you experience business leaders facing and how do you recommend they get past those?

Stefanos Loukakos: So there are a lot of challenges when you deploy AI. I’ll start with the easy ones. Hallucinations, what happens when a company might sell a product that doesn’t exist or a product that they don’t have on their inventory? So we need to solve for hallucinations. Out-of-scope conversation, which is really important and it aligns with brand identity. What if an AI of a big retailer is having a conversation with a customer and the customer start asking questions about the war in Ukraine or the 2024 elections and things like that. How do the AI will respond?
How would that fit with a brand identity? Another big challenge is integrations. So for the AI to really work, it needs to be integrated with the business and the retailer’s backend systems, product catalog, e-commerce engine, CRM, et cetera, et cetera. Why? Because it needs in real time the correct information to be able to have meaningful conversation with the end customer. The other really important challenge is vertical specific sales. If you train an AI to sell, I don’t know, fashion clothes and things like that, that’s very different from an AI that sells cars or an AI that sells healthcare products. As I said, again, brand identity is a big one, a big retailer and a luxury brand would have totally different ways of communicating with their customers. So the AI needs to capture those. And another big barrier that is not, again, simple to solve is language.
Today, the majority of NLP, natural language models, et cetera, are in English. Yet a lot of customers and AI is deployed in a lot of different markets. For example, Connectly we are deploying our AI solution in Indonesia, in Brazil, in Colombia, in UAE like in the Middle East. Totally different languages, very different the way people communicate with the brands. So these are a couple of challenges that businesses have to look at, and these are not easy problems. It takes a lot of data, a lot of training models to solve this. But can I just say one more point? I believe, and again don’t count my opinion as an expert, but I believe that there’s not going to be one AI that’s going to solve for everything. There are going to be vertical specific and use case specific AIs that are going to solve specific problems.
So AI for retailers for sales, AI for healthcare, AI for hospitality, AI for you name it, for banking or for finance. So I believe the more you focus, the more you focus means, the more you get data that are very specific for this vertical end use case, the better you can train your models to solve actual problems and issues for the consumer.

Amanda Razani: So with this big transition in focus, do you see a lot of companies forming divisions just focused on AI implementation and just focused on going digital?

Stefanos Loukakos: So everybody’s talking about it, as you probably know and that’s probably why we’re doing this podcast, but we haven’t seen a lot of solutions out there that are really applications of AI. Today I believe the biggest, and this is from what I read right, the biggest investment is on the infrastructure side to build these models, to create the infrastructure for this model to be able to train AI, et cetera, et cetera. But we haven’t seen a lot of actual solutions that are helping specific industries or vertical or use cases being deployed. We have one of them, it works really well. It needs a lot of work to become even better, but we believe that in the future it’s going to be a combination of both businesses using their own building, their own models with their large businesses, but the majority of medium-sized or smaller businesses are probably going to use all the self-solutions for the specific vertical and use case they want to solve for.

Amanda Razani: This technology is advancing so rapidly and as more people do conduct business online, more people expect a certain higher level of customer experience. So looking toward the future, what do you think is next for business leaders? What do they need to focus on next when it comes to that customer experience?

Stefanos Loukakos: Great. So we have a lot of stuff in mind, and again, I’m going to mention a couple of stuff that we believe is the future of how AI is going to work. And I’m going to talk a little bit more about our field, which is we build conversational assistants that help businesses sell products, e-commerce and retailers sell their products. So what do we believe is going to happen in the future? More advanced personalization, it is going to be way more personalized messages, personalized questions by knowing the customer better. This will happen with the help of integration. So I would know, for example, what you have bought in the past, if you are an existing customer and an old customer, what colors you like, et cetera, et cetera. So I would be able to have a conversation with you that was way more personal as if I know you.
Okay. The second thing is today we see a lot of chat-based assistants. I believe we’re going to see a lot of more voice-activated assistants. And these are not going to be like talking to a robot, it’s probably going to be talking to an actual person, and you shouldn’t be able to understand the difference. Omnichannel is another thing that is going to continue to grow. So you’re going to have assistants on chat, on the website, on everywhere. And the businesses need to have conversation with the customers on any channel, not only online. Predictive analytics. So what I mentioned before, businesses are going to use analytics to really understand and forecast what customers would like. So a very simple example, I’m a marathon runner. I buy shoes every six months and I change my shoes every six months. And AI should kind of know that. And just before my old shoes go bad, they should come up to me and show me how to buy the best new shoes in order to continue running and things like that.
And I believe another big part of the future of AI would be the AR and VR integrations. We’re going to see immersive shopping experiences for retailers where they’re going to be able to talk to an AI and even, not touch but feel the products, et cetera, et cetera. So it is a very exciting future. A lot of people are thinking about these problems, very hard problems to solve. But again, it’s one step at a time. The first thing we need to do is get out of this hype of AI can do everything to actually do stuff for businesses and help them increase sales, increase conversions and things like that. But it’s a very exciting future anyway.

Amanda Razani: Absolutely. And there are some great use case scenarios to consider. So what’s one key takeaway you would like to leave our audience with today?

Stefanos Loukakos: One key takeaway would be that AI is definitely going to change how businesses operate. So businesses should try to use new tools, but they should really evaluate the tools based on results and not on hype. They should only use tools that are going to actually help their business grow. And they should experiment because the solutions out there are all new and we need to really iterate before we find out exactly what works and what doesn’t.

Amanda Razani: Well, thank you so much for coming on and sharing your insights. It was a great conversation.

Stefanos Loukakos: Thank you so much.

Amanda Razani: Look forward to speaking with you again soon.

Stefanos Loukakos: Thank you.