Digital and Voice Channels Are Converging

In this episode, we chat with Endre Davids, founder and CRO of two companies,
Conversational (Conversational.nl) and Seamly (Seamly.ai).

We will cover:

– The story of Conversational and Seamly, and how these two brands coexist
– The evolution of text-to-speech and speech-to-text technology
– The role of artificial intelligence in the near future
– Their feedback on working with Inbenta

Listen to the full episode or read the transcription below. 💪

Jordi Torras:
Hello, and welcome everybody to the next chapter of the Future of Customer Service. We are really honored today to have here with us Endre David. He’s a founder and CRO of two companies, actually. And he’s going to talk about that, Conversational, you can find Conversational at Conversational.nl and Seamly, Seamly.ai.
I am in Santa Cruz, California. Andre is in the Netherlands, not far away from Utrecht right now. So we are having this conversation with all the bits and bites going under the Atlantic Ocean, I guess. So that’s pretty cool about technology. So Endre, thank you so much and welcome to the show.

Endre Davids:
Thank you for having me.

Jordi Torras:
Absolutely. So could you please introduce yourself? Who are you? That will be awesome.

Endre Davids:
Yeah. I’m well, I’m Andre David, working and living in the Netherlands. Originally half from South America, my mother is a Chilean, raised in Spanish as well, but my Spanish has become too rusty to have this conversation in Spanish, unfortunately. But I live 20 kilometers from Utrecht with my wife and my two kids, nine and seven. And currently, it’s five o’clock, well, quarter a past five on Friday afternoon.

Jordi Torras:
Nice. Eight o’clock in the morning, that’s amazing. So could you let us know about the company, Conversational and also the story of Seamly and how these two brands coexist?

Endre Davids:
Yeah. Almost five years ago we started off with Conversational as a consultancy firm to help organizations in the adoption of self-service platforms. And we did that for quite some big clients, among the biggest telecom clients in the Netherlands and big energy companies. And during that process, during those implementations, quite often, we also suggested how they could enhance their either interfaces or even integrations with the chatbot that was just introduced.

But most of the time, our clients struggled with the actual creation of those suggested optimizations. So what we initially ended up doing is that, first, we build those integrations, well, as a fully custom work for our clients. And we did that two times, well, two other times as well and then we thought, “Hey, maybe we should productize this.” And then we developed and introduced our own conversational presentation and integration layer, which is called Seamly.ai.

Jordi Torras:
Wow. So that’s a story about productization, going and saying, “Okay. I’m building this custom work for customer number one, then customer number two, customer number three.” And then you say, “Wait a minute, I’m doing the same, not the same, but kind of the same. So let’s put that together into a platform that actually will be able to reproduce and scale.” That’s lovely.

And the story sounds familiar to me because we at Inbenta as well, we started as a consulting business doing custom work and kind of a similar thing like, “Hey, we are doing the same thing over and over again, let’s have our computational linguists work together and create a superior engine that will work as for natural language processing.” So that’s, that’s amazing.

So if you take Conversational, I’m assuming that Conversational still is doing consulting and working for customers as professional services, right? So what is the typical problem that you guys solve for your customers?

Endre Davids:
Well, the biggest problem and the biggest issue that our customers are facing is that, in the last recent years, most of our clients have bought quite some new technologies and introduced quite a lot of new technologies. But they’re hardly able to actually make the biggest benefit of the best use of those platforms that have been introduced. So that’s one element, to make sure that all these solutions that are in place, that it can actually work better together.

The second aspect that we help our clients with, is that we offer great user experiences. Fully in the look and feel, fully in Conversational interfaces that are fully in the look and feel of clients like DHL, for instance, or our banks, or in insurance companies that we have as clients. And thirdly, if we have integrated so many solutions, and if we have great Conversational interfaces. What we then do is that from those conversational interfaces we make sure that we offer single and seamless conversations. Even though a user might switch from one system to the other, the user isn’t bothered with that. And all these systems work seamlessly together.

And the last reason and solution that we help our clients with, is that we actually help them speed up their innovation process, and not create the necessity for them that each, well, let’s say feature or innovation should be specified, built, and maintained by themselves. But, well, I can say that we work for the biggest and the best North European cases. So our clients, they actually get the benefit of the being in the front run are with that as well.

Jordi Torras:
Nice. So all these Conversational AI, it’s great, you can buy, there are dozens and probably hundreds of different platforms to choose from, but there is a common fitter here, it’s like they say, it’s like, “Okay, this is the tool, so you, Mr. Customer, now can build your dialogue or your conversational strategy.” And that’s something that is peculiar it’s not just writing just content. It’s not programming like in a programming language, it’s something specific.

So every customer that I know needs help on that and trying to understand the nuances behind that. And sometimes, Endre, they ask me, “Hey, what’s the kind of profile that we need to hire in order to implement all that? What I’m going to say is like, “Well, call Conversational, they will do it for you.” But then, how would you describe the skill of the professionals that build Conversational’s AI?

Endre Davids:
Well, if you look at Conversational copywriters, then I think that there are a couple of elements that have to be included in that role. Of course, you need vision on how to create an identity and how to use that identity in conversational copywriting to the mix. Secondly, I think that there is a big challenge that, you also need to have the ability to transform, well, let’s say, company information or procedures or processes into bits and pieces that are easy to understand. And can even be used in a dialogue way of interaction, rather than just a lot of content placed on one screen.

And the second or the third element that I see as a requirement is that you need to have colleagues that are really enthusiastic about user experience. And what happens if a user says A, does he get the right response? What happens if the user in the flow wishes to go for another flow or submits input that wasn’t expected? So I think those are the three elements. It’s more like, well, creativeness, well, being able to use company data and procedures and having a really good eye for details.

Jordi Torras:
Wow. I think you nailed it, that’s exactly the kind of skills needed for that. We call these roles sometimes botmaster. We use the time, I don’t know, I think it was the ’90s already, when they would call the webmaster, who was the person who would create the website, now, it’s entire teams or companies. Yeah. But we call it botmaster, and it’s amazing the kind of a specialization that you guys have been building around that.

I think poking around your website, conversational.nl is amazing. And I’ve seen, then you have an important aspect of that is working with partners and partnerships. So I would love to know what is your vision about these technology partners that you work with?

Endre Davids:
Yeah, yeah. If you look at those partnerships, mainly those partnerships are all focused on technology companies that really want to exceed in what they’re doing. So if you look at your organization, for instance, trying to build the best Conversational platform and achieving quite a lot in that. But from that side, also making also realigning yourself in that is your expertise.

And just like that, there are tons of really great live chat applications. And what we’ve seen is that, where there’s a gap that can be bridged for our clients, is that if we bring those two together in really great interfaces, for instance. So all our partnerships are based on the fact, of working with organizations that are focusing on doing the best that they’re offering and realizing the best possible results in that.

And based on that, we work quite a lot with, well, of course, with you, but also with platforms like Genesis, or Avia, or Life Force, or even Salesforce or Microsoft Dynamics, for instance, or well, customer feedback solutions. And these examples are quite different examples, but then we always make sure that if your user starts a conversation, from that one conversation, you can go from a chatbot conversation to a live chat agent, or even possibly change your content details in the back end of a company without having to, well, switch in screens or in journeys and start over again.

Jordi Torras:
Oh, yes. And one of the things that I’ve seen you guys doing is, for example, integrating Conversational AI with automatic translation. And you guys did this workshop the other day was amazing. Okay. Now, you implement in one language, and then using automatic translation, you actually have your Conversational AI with multiple languages.

Endre Davids:
I think we do 107 at this moment, different languages.

Jordi Torras:
That’s amazing. By combining, let’s say, different AI, and then productizing the whole solution under Seamly, I think is a brilliant idea. One of the things that we have seen is more and more digital channels converging, and converging with voice base. So we have seen the voice world deal with call centers, IVRs, and these kinds of interfaces. How do you think these digital channels and voice channels are converging? What is your experience in that?

Endre Davids:
Yeah. I think they’re converging quite rapidly and also in a quite successful manner. Well, even we just introduced a voice bot based on the Inbenta platform, a couple of, well, two weeks ago for an international client. Directly, by the way, in four different languages indeed. And my view on that market is that, text-to-speech and speech-to-text, wasn’t that great, let’s say, 10 years ago. But in the last couple of years, text-to-speech has become extremely natural.

So there are quite some good voices available, and speech-to-text services have become pretty spot on as well. So if we look at what happened in, let’s say, maybe 10 years or five years ago, there was also a necessity where things needed to be optimized and needed to be, let’s say, separately managed and continuously needed to be optimized for that.

Well, if we look at the current available text-to-speech and speech-to-text technology, and if we add that to the Conversational platforms that we supercharge, then you have full flexibility and fully reusability of what you already offer on your website with the chatbot or recognition and the answers and reuse that maximally on the phone. And that creates a lot of content, and helps organizations tremendously in flexibility and in the creation and optimization of content.

Jordi Torras:
Yeah. Absolutely. So it looks like at some point voice is going to be considered yet another digital channel, because it’s being digitalized very quickly by services. As you say, speech-to-text technology is improving and speed is everything. When we humans are talking to each other in our own natural languages, English, Spanish, Dutch, whatever, we are really fast at understanding each other, if you speak the language of course. And machines, well, still not there, but they’re getting better very, very quickly. I really agree with that. Amazing.

Endre Davids:
Yeah. And I think what the essential part is, and what we always aim for is that we look at what we can digitize in a conversation and where the human can actually, or the human agent can actually add a lot of value. And make sure that even in a phone call we digitize the, let’s say, boring part of the conversation where you need identification of the user, ask first a question. Maybe the user should even go through, a rather old classic IVR menu, like in, “Press one for this. Press two.” And if you have to wait for option number six, then-

Jordi Torras:
And you don’t remember when it was option number two?

Endre Davids:
Absolutely. Then you’re either going to choose the different option, the wrong option, or you’re going to be irritated already. So that’s why we also make sure that the digital part of the conversation seamlessly works together with the agent and the agent can pick up the conversation where the preparation has left enough for him or her.

Jordi Torras:
And sometimes I’ve been in, maybe not lately, but at some point I had the conversation like, “Hey, what about AI? Is AI taking away jobs and yada, yada,” this is a story about artificial intelligence. And what I say is that AI is here to free humans from boring jobs. And I like to use the word boring because that’s exactly what AI does, is take away what is boring and free humankind to do what is interesting, and there’s creativity and there’s value in human communications.

So if you’re able to have AI approaching on all that, that’s essentially a value that you are adding and you improve the scalability of businesses. And what do you think is the role of artificial intelligence and all that, and how do you see this trend? How is that evolving?

Endre Davids:
Yeah. I think we would hardly be able to live in a world without AI, which already starts with the smaller touchpoints that you have with your, let’s say, smartphone, for instance, or traffic jam predictions. So I think there are a lot of things that we’ve become so used to that we would struggle quite a bit if we have to drop all that. And I think AI in all its different shapes and sizes has become such an important element of how users also interact with technology.

That AI, in some cases, also helps the end-user to maximally use that technology available. And from an organizational point of view, I think that AI, especially reduces, of course, costs, but also improves staff retention and customer loyalty. And with that, increases revenue and even customer satisfaction. And all those elements, if you choose not to supercharge them with AI, then you’re going to have quite some struggles in the markets compared to your competitors.

Jordi Torras:
That’s right. And when I was in college, I got specialized in what it was back then, artificial intelligence, a while back. And one of the things that we learned is that when you have this technology which is in the space of artificial intelligence, and then it works, and then everybody uses it, nobody thinks that is artificial intelligence anymore. Because I remember back then, one of the artificial intelligence problems that we were dealing with was finding the optimal path between different route alternatives.

And today we take Apple Maps or Google Maps, and we take it for granted. But that used to be called artificial intelligence, and indeed they are using a lot of machine learning and artificial intelligence to predict with extreme accuracy, “When are you going to arrive?” So that’s pretty amazing. And yeah, I like that, we cannot live already without artificial intelligence. So totally, it’s here to stay.

Endre Davids:
Even if you look at new technology that has been adopted, well, in the intro, I told you that I have two kids. And I think it was two or three years ago that my daughter was in the, well, she started practicing math or calculations. And there was one day that she asked me, “With how many people are we going to go on our holiday?” And I asked her, “Well, what do you think?”

And she said, “I think we’re going to go with six.” I said, “With six, well, how do you get to six? Because how many are we in our family?” She said, “Four.” And I said, “But four is not six, right?” “No, no. But we have a dog.” I said, “Okay. But if we add one to four, what do we get? Well, we’re at five. But five isn’t six, right?” “No, no.” And then I asked her, “Who’s the sixth person then in our family, Google?”

Jordi Torras:
Oh boy.

Endre Davids:
She expected that our Google assistant, of course, would join us on our holiday as well.

Jordi Torras:
Oh boy, that’s cute and terrifying at the same time.

Endre Davids:
Exactly. Yeah, yeah. It worried me as well from one side.

Jordi Torras:
Oh, that’s amazing. So we know each other, right? And we’ve been working with Seamly together with Inbenta and different customers. Actually, we met because of a common customer, which is amazing. When the customers start having these building blocks of technology. And you can be as cruel as you want, but how has it been working with Inbenta and all these projects?

Endre Davids:
Yeah. There are three things that really struck us, and that really we’re impressed with. The first one is, the way your engine automatically prepares content and is really, well, pretty impressive in how certain answers are already prepared. And secondly, is the way that we can cluster content and, well, use that for intent recognition. So for instance, in the voice bot example that we develop on your platform, if the user on the phone shares what the question is.

Then based on the organization of content in Inbenta we can already derive the fact that maybe it’s a question about, well, about a product, and even a name, the specific product itself, that it concerns. So that’s really impressive. And lastly, to be honest, no matter how good technology is, a big differentiator is always the human side of working together.

And what we really like is the way we can share our challenges or ideas and even go, well, get to conclusions or get to suggestions or solutions generally. And sometimes that’s a solution that can already be used if we organize things differently. Or sometimes it’s a solution that is being put on your roadmap. Well, the human side is something that we’re really enjoying.

Jordi Torras:
Oh, thank you so much. I really appreciate it. And as I said, we’ve seen applications that you have built with Inbenta and other technologies. And it’s like, “What? That looks amazing.” And especially the touch that you give to this user experience is amazing. And yeah, it’s such a pleasure working with you guys. So let’s talk about the future, how do you think the future of customer service will be? Which, by the way, is the name of this podcast. How is your vision about where things are going to be in the next five to 15 years?

Endre Davids:
Yeah. Well, five to fix 15, that’s quite a challenge. But if we look at the next, let’s say, two or three years, I think that, especially from the user side, further increase of cell service will be realized. If we look at current data, then about two-third of all the customers that have a question, actually try to find that answer to that question themselves.

But in over 60%, they can’t find their answer or what they’re looking for, and still need to reach out to a company. And then maybe call or the company picks up the telephone call and thinks, “Oh, this is the first contact point, let’s try to fix it in a one-stop-shop.” A first-time fix is something that companies use a lot. But I think what companies aren’t realizing good enough is that, that actual phone call in two-third of all the, well, let’s say 60% of all the cases, is actually already a second attempt.

So further increase of self-service is something that we’re expecting. Also with that conversational-based technology, will help to facilitate that. Where a user submits a question, and then actually that question is being interpreted and provided with one answer rather than a list of 10 or tens, or even hundreds of articles where the answer may be included. But I think that’s not going to be accepted by end-users in a couple of years anymore.

Seamless handovers to agents, if still assistance is required, that is going to be a necessity. If you don’t have that as an organization, I don’t think you’ll be considered, well, let’s say, as a valuable organization, or even as an organization that respects its clients. And those seamless handovers to either live chat, they can also become cross-channel.

So if I’m having a conversation with a virtual assistant, and then maybe I want something emphasized or I have a question that I can even switch to either a phone call, where the agent on the phone already knows what we’ve shared digitally with the virtual assistant. So that’s something that I can see, that I expect to be coming. Well, from the business side, shortage of personnel is going to be a challenge where.

What we expect is that a lot of attention is going to be spent on, “How can we facilitate and optimize the journey for the agent? Where can we remove the, well, the boring parts of the job. But also the parts that can be automated and making sure that working in, let’s say, customer service specific department is better than working as a customer service department of your competitor,” for instance.

And with the last two years, of course, that we went through, remote working has boosted further potential for internationalization of customer care as well. Where indeed language barriers can fully be evaporated by using automated translational services that we do for our clients.

Jordi Torras:
Oh, that would be amazing.

Endre Davids:
Yeah. So those are some examples of developments indeed, that we see happening with our-

Jordi Torras:
Totally. That is great. Listen, I would stay here for hours talking to you, because I believe we share this passion about AI, and user experience, but we have some limitations on time here on our podcast. So what I would like to ask you is, for the audience that is listening to this now, how could they contact you? How could they contact Andre David? That might sound bad. So how can the audience contact you?

Endre Davids:
Well, of course, I can be fine on LinkedIn. Probably my name will be in the screen, there aren’t too many Endre Davids in the world. So that can’t be that much of a challenge. And of course, [email protected] And then Endre is spelled with E-N-D-R-E. So in Dutch it’s Endre, but in Spanish it’s Andreas, and in English it’s Andre. So I’m fully okay with that.

Jordi Torras:
All right. Yeah.

Endre Davids:
But it’s spelled-

Jordi Torras:
It’s an international name. Yep.

Endre Davids:
Yeah. Absolutely.

Jordi Torras:
In my mother tongue, I’m Catalan, originally from Barcelona. So in our language it’s Andrew. Go figure it. So there’s Andreas in Spanish, Andrew, and I think similar names exist all around the European languages. All right. Thank you so much for being here. Really a pleasure. Let’s keep in touch.

And for the audience of this podcast, very soon, we’re going to be having more interesting, we’re going to try to keep this as interesting as with Endre, we’ll see. But we are very excited about having Endre, today, here with us. Thank you.

Endre Davids:
You’re welcome.

Jordi Torras:
And Endre, see you on the next podcast. Bye-bye.

Endre Davids:
Bye-bye.