LexiQA currently supports 150 locales. They started with four. What was the journey from 4 to 150 locales? Find out in this interview with 2 out of 3 co-founders of LexiQA – Yannis Evangelou and Vassilis Korkas.
Welcome to a new episode of our series “How I Started”, where we will focus on founders of different localization businesses.
In this episode, join us as Vassilis and Yannis share about:
- The idea behind lexiQA
- Growing up in Greece, teaching at a Uni, their shared principles
- Why linguists like working with them
- How to be more responsible with your startup
- Why buyers develop their own platforms
Andrej Zito
Let’s start very simple. Where are you guys from?
Vassilis Korkas
We’re both from Greece. In terms of citizenship right now, though, we are located in different parts of the worlds. When we started the company, I was based in England. I was living there since 98. And that’s how I bumped into Yanis. In my previous career, when I was teaching translation at University of Surrey, and Yanis, now, he will tell you his own journey from Greece, to Berlin. I mean, now, as well, in the last four years, I’ve been living in Riga, Latvia. So we’ve been all over the place. But this is like, how flexco has been, since the beginning completely decentralised. And we were doing work from home when it wasn’t trendy.
Andrej Zito
Alright. We’ll get to the part where you to actually meet a little bit later on. Maybe give us a context, because I know that we were joking about this during our first call. Can you give us maybe some sense of how how distant you are? When it comes to the H? Liquid is the age difference between you
Vassilis Korkas
know, that big? No, young
Yannis Evangelou
person, the three I’m called to save? Okay.
Vassilis Korkas
Because my interest is somewhere in between. So I think oh, no, it’s
Andrej Zito
okay. Yeah. Because I think that when you were first telling me that, I don’t know, like, Basilia was the university teacher, I thought you were like, I don’t know, like a fresh graduate or something like that. Like, there was like a huge gap. So no, okay. So you are almost the
Vassilis Korkas
things. When I started teaching, I was also very young. Basically, I started the year after I finished my master’s. So I was 26, at the time, and I met young, when I was I don’t know, in my early 30s, I guess. So it was all. And you got this all the time. I mean, I remember at the time, we were getting groups of seven ad students and all the time, you will get maybe four or five students in that cohorts, which were older than ours. That was perfectly normal, right? It wasn’t just fresh graduates out of language school, you got all sorts. And there were people who would come in with experience in translation, not a lot necessarily, but especially the the older students back then they were either looking to have a career change, let’s say, or they were already working in translation for a few years, but they wanted to get a kind of certification, something that would help them develop their careers a bit more. So, you know, different people have different motivations to pursue an MA. And they still do, I mean, it’s still the case. It’s just that now all the circumstances in higher education are very different in the UK, especially the numbers of students who have been studying languages at lower higher education, let’s say in high school, or anything like that. They’re far fewer now than they were before. They have been declining for years, in fact, so now you get into a situation where in the UK, you can hardly get enough translators into English, the same problem then spills out in the EU institutions, for example. In a way, it’s good, that Brexit happened, because now only Irish translators work in the group, so gonna have to worry about it. But it’s something that does have an effect on the industry as a whole, because that kind of creates a bottleneck in the supply of new translators in the fields. And that means fewer people can then start thinking about specialising into technical translation, or localization or engineering or things like that. And that’s where a lot more of the math of the demand in the industry will come in the future. It’s not going to be just traditional, let’s say translations. Binary, right? That’s a completely different conversation, of course,
Andrej Zito
right? I was just gonna say that. Okay, I’m gonna play along and continue asking the follow up questions. But if you say it’s a different topic, then okay, maybe let’s get it right.
Vassilis Korkas
Well, let’s say I don’t want to go too much into things that I haven’t been in touch with for quite a long time. Now. The truth is, I left academia like eight years ago, almost so it would be hard for For me to claim that I know exactly what’s going on right now. But my experience was that kind of decline that had started before I left is that now there is a much bigger shortage than before, in terms of new people coming into the industry.
Andrej Zito
Okay, so I still have to ask, since we’re still talking about it. So even even if you when you were in the, in the, in the part of the academic world, like, why did you think this decline was happening?
Vassilis Korkas
Well, in the UK, specifically, the bigger reason had to do with a systemic problem in the UK, which started, I’d say, about 1520 years ago, when they start promoting STEM subjects in lower education, more than humanities, social sciences, etc, which meant that a lot here, younger students were interested in languages at all, if they were lucky, they would get one foreign language to be taught as an option in many schools. And that wouldn’t necessarily be even promoted actively, right. So most of them would need to get all their A levels, in maths and physics, etc. But whatever foreign language they had in their, in their arsenal, it wasn’t necessarily pursued after they finished with lower education. So when they had to apply for courses in university, they might get, let’s say, a mix degree where they get to study in business and Spanish, for example, or something like that, but not necessarily, let’s say language and translation, translation would come, usually as a kind of an after assaults and postgraduate education, at least in the UK, for the majority of cases, in other countries, things might be different. And I’m sure they are. In Greece, for example, it’s almost compulsory to have to language to foreign languages in high school, for example, one is compulsory, and then you get to choose one amongst three other choices, let’s say. So the idea of learning a foreign language is something that is much more promoted. And it’s not frowned upon. It’s encouraged. So the UK generally has a different market, both in terms of education, but also in terms of employment later meant that they wanted to put emphasis on other industries that they wanted to develop further. The other issue that also played the role had to do with the fact that both in education and industry, the UK as part of the European Union, then could attract a lot of talent from other countries. So they didn’t feel that they had to necessarily develop their own resources in the long run, they could make up for the loss of talent by recruiting talent from other countries. And that was the case, that is still the case, for many industries, because there are people now who work in many of these industries that have been living there for more than 10 years. They’re permanent residents, they have the right to live there. So there’s no problem. But for other sectors where this mobility didn’t exist, it’s becoming more problematic to make up for the loss in talent. So now, I think this problem is gonna be even further exacerbated. It’s now that the UK is not part of the EU, they will not have the kind of resourcing availability anymore. And we will probably see LSPs struggling to hire people locally in the UK. And they will have to make up for those losses by recruiting people remotely. It’s not a bad thing, of course, because now you have a broader horizon. But finding the right fit for your company might be more challenging, because exactly. You don’t see them at the office. Right? You don’t know who they are basically.
Andrej Zito
One thing that I’m curious curious about so the way I understood is that the systematic reason was that, I don’t know, let’s say the government or the education system started promoting more STEM subjects versus translation. I have a very weird parallel that maybe I’ll share later on. But even if, let’s say students were exposed to, I don’t know, translation studies and STEM at the same time, do you think that people would actually choose translation, or do you think like, ultimately, it’s been Based on what people like, like, do you think like more exposure to something could tip the balance a little bit, or,
Vassilis Korkas
well, less exposure can not be a bad thing. So if we look at it the other way around, the more exposure that they would get to this combination of STEM plus languages, let’s say, not even translation, the chances would be higher that they might see technical translation, let’s say as a potential career path. And these days, we wouldn’t even be talking about technical translation, most of the time, we will be talking about language engineering, natural language processing, machine translation, all those things that are basically exactly the interface between language studies. And computing, for example, or hardcore localization engineer, at least, in either case, there are roles today in the industry that didn’t even exist, like three or five years ago. So even if I tried to compare to what was available in the industry, when I was teaching, I wouldn’t be able to tell you that there is an overlap, because there are so many new things out there right now. And academia is always slower to react. Because as an institution, it is slower. From a regulatory point of view, you know, you have to jump through a lot of hoops to create a new programme, a new study programme, and to get your new module descriptions to get new tutors to find the right people to teach the subjects and make sure you have enough students to run the course. Because if you don’t, then you don’t run it. There are all sorts of challenges, even for well established universities, these are challenges. And now that’s they don’t have the same, let’s say funding source by recruiting more students from the EU, who used to have easier access to these institutions. Now they have to rely on other methods, let’s say in order to attract more students, so they might get more students from Asia who have to pay higher fees. Or they have to diversify the number of modules in the study programmes, which might be optional or compulsory, so that they can attract students from other parts of the university to study modules in the language programmes, let’s say, and they pick and mix. That way they have bigger numbers in their classes. But you know, these are institutional challenges for universities. Generally, I would say not just in the UK, but the UK have this additional layer of complexity because of the lack of Native students who would be willing to further their language studies to go to high level, right or not. So we could talk about it for a very long time. But you know, as I said, I’m a bit reluctant to go into more details on that topic, in the sense that I’ve been away from the, from the fields in a while,
Andrej Zito
we also have a second guest here.
Andrej Zito
Let’s go back, let’s go back to the early days, because I like talking about these topics. So maybe you can you can start with this one. How do you remember your childhood? Growing up in Greece? Yeah. Like the first thing that comes to your mind is it the positive experience.
Yannis Evangelou
It is a positive experience, which had to do with a lot of play and a low load of studying. So back then, we didn’t have the chance to learn languages at school. So we had to go to private schools, private language schools, after our classes. That’s how it worked. So I started learning English and French, very early. So English when I was starting up and French when I was eight, and I did finish my studies in both languages back then when I was almost 16. But these were courses that we’ll have to take in private after classes. So you know, you wanted to have an extra activities like music or football or whatever. It all amounted to very steady schedule. So at some point, when you were preparing for the university, let’s say from 16 to 18, everything had already become too much. There was a lot too much weight on your, on your shoulders, and it’s not It’s very easy to handle a situation like that, at that age, especially, you know, when most of our parents were not able to understand what was going on. Now, they just wanted us to join the university, you know, get their degree and get a job. And that’s it. Like, what it used to be. So, obviously, back then, people didn’t know anything about translation, I had the opportunity to have a relative who used to be like half of who is half Australian. And I did some classes with her when I was 15. And said, you know, what, if you get that degree, then he then there’s another degree, which is a translation degree. I said, So what’s that, and it was like an extra translation degree from English into Greek and Greek into English. Which, quite frankly, what I succeeded back then, but I’m, I’m pretty confident I would fail. Now, if I tried the same thing. Like I had, much back then, that I succeeded. That was 95. But now, yeah, I will come down into failed anyway. And the system in Greece doesn’t really is not really ideal. So if you want to become a translator, you have and you want to study translation at the university, you have to go through a specific syllabus, you have to go there for days. It used to be like that. Now it has changed a bit, but nothing has really changed in the core. So at the end of the day, I studied mathematics at the University, I dropped after five years before I get my degree. So five years later, I dropped mathematics. And I switched into translation, there was the British Institute. Back then it was like a private institution started their translation from English into Greek and Greek into English for two years. And the British Council Institute, sorry, British Council. And I also lead the French into Greek and Greek and French degree, similar institution in my second year, so then I applied will have the chance to apply for a master’s in the UK. And was a few months later, that was fine. I landed in Surrey, with my huge suitcase, which included the desktop. Even academics, it was my first slide, I didn’t know what people carry with them in their flights. So I just put everything. And I remember I remember the gearbox, and I said, Sir, this is 57 kilos, and the limit is 25. Sorry, I was born in a village. I don’t know how this applies. Okay. Next time, it’s fine. The problem was, once I landed in, sorry, I could not open the suitcase. So an issue there, anyway. And this is how I got to know Bill Phillips, who was my one of my professors, at the Masters, and so on, if I did masters, I applied for a PhD and looked for a job in sorry, Phyllis also helped me get that first job back in 2006, as a QA specialist at the local LSP. And now this part of the story does mustelids difficulty students?
Andrej Zito
I think I think if I if I if I can hit a pause, because you guys really like going to your to your school area in the university. And when you when you met, I actually took a note here, like the first thing that you said that you had to study the language of the private school, does that mean that you had to be from a certain family, let’s say income wise to actually even go there.
Yannis Evangelou
It was like pretty much everybody was attending that kind of private school. They were like, cheap private schools. So I didn’t have private classes with a teacher, the ones who came from rich families, had private classes with an expensive teacher of English
Vassilis Korkas
class, the tutor would come at your place, he would sit down for an hour or two and have a lesson like for a couple of times a week, let’s say. But otherwise, you would do it in in a school, but it was a language school. It was just for that for nothing else.
Andrej Zito
And you say and you’re saying this is still going on? Oh,
Vassilis Korkas
yeah. Greece. Really? Yes. Yeah. It’s very popular. So.
Andrej Zito
So is that the main way how Greek youth learns new languages through this private
Vassilis Korkas
Well, if you thinking of the formal system, then yes, otherwise they learn English, let’s say through watching television, or listening to music. Everything that works for everyone else works the same way in Greece too.
Andrej Zito
Can you explain to me why is it not integrated into the let’s say the main education is
Vassilis Korkas
more integrated now than it used to be. But back then, let’s say language learning was more like an optional class, rather than a compulsory one. So you had the choice not to take it. Whereas nowadays, it’s considered absolutely essential. But also the fact that the level of tuition, the kind of quality of tuition that you would get in a public school was not as strong was not as good. So a lot of people would go to the private schools to compensates for whatever they didn’t get into public school.
Yannis Evangelou
It was also that everything was starting kind of late. Meaning that when I was asking my parents, I said, You know what, I want to learn English, and I was five and a half. We don’t know what that is. But anyway, let’s do it. But otherwise, my next option would be to start at the age of 13, at school and do one or two hours max of English a week, meaning that, you know, at 18, you were still like, I ate you ate, and so on. The same level. So now, I think they started learning. Earlier,
Vassilis Korkas
yeah, even I think B class, let’s say, a second year at school, you would start learning the foreign language. It’s been like that for years. Now.
Andrej Zito
I think that that’s very interesting. Danny, what you just said, because my question was, at first, I was like, why did you decide to study languages, optionally, at such an early age, and then I was like, wait, you are just so young. So your parents probably forced you. But now you just said that you actually went to your parents to ask about it. So can you explain to me why does a six year old want to learn languages? Like why? Why did you feel the need to know English?
Yannis Evangelou
Good question. Because there was no there was nothing in my family in my broader family environment, which would kind of help me understand what what languages I mean, there was nothing which could, which would even give me a sign with regards to languages or translation or language learning whatever, or even with regards to books, like there was no single book. At home, there was no book, right. Delish, or engine No, in any, any language. And about the, I’m talking about any language. And I still remember it was like, when, when I finished my first year at the primary school, my teacher back then who I have to say, is still alive. Now she’s almost 90 years old, not 85 or something. And we still call each other once a year, see weird?
Andrej Zito
Did you have a crush on her when he or she
Yannis Evangelou
will characters will cut this editing anyway, and she told me my mother, she gave her a list of books, she told my mother, you know, what this guy is really interested in what we’re doing here is not like the other, like the rest of the people who check out of the window all the time, and so on. He’s really focused, some giving you a list of books by him. My mother, God bless you both with the books, and then I need more books and so on. This is how everything started. So this is why when I started mathematics, which I love, later on, I was not really motivated to get my degree in math, and then do a related job just because, you know, that was a shift away like a huge pivot away from what I was wanting to do. I wanted to do something which would combine languages with computers is where I loved. I’ve always loved languages, literature, computers and math. can I combine all the two all the all four cert up to the or can I combine all four? Well, back in the day, it seems that I couldn’t be able to do that. Not only Chem like NLP in a way is what is probably the way to combine all four. But yeah, it’s not something that I hope 20 years ago, back then it was like, Okay, I’m dropping math to get into translation and become interested become translator. And we’ll take it from there. So it was a failure. Like I didn’t get it was a huge failure, right? I didn’t get my degree in math. So what am I doing? If I didn’t switch the translation, immediately a switch to translate? I should speak kickboxing for a year as a translation a year later, like, Okay, now I’m mature enough, after getting beaten 1000 times, I’m mature enough. moved to an office job. Yes. Work, it works. Everyone watches this podcast to do the same.
Andrej Zito
I’m not sure if I zoned out or something. But I’m still not sure if I if I understand like, Why does a young person want to learn languages? Like what drove drove you to it? Was it curiosity?
Yannis Evangelou
So it was what I think because I’m not sure like, I’ve tried to decode this. It’s a great question. Because it for me has been like a psychological question. I’ve been trying to dive into the source of this for a long time. And it was, it was like, my father moved to Germany before I was born. Right. But he moved back to Greece a couple of years later. And I don’t know why I had a crush. With regards to Germany. I like I always wanted to live in Germany, when I was, I think had to do with goodwill and so on, and what we were watching on television, and so on, we had the thing, like all of our neighbours, we had three neighbours, and all of them used to live in Germany. So they were bringing videotapes VHS, but everything was in German. So it was Jackie Chan in German, you know, it was mostly about karate, and Western, right.
Yannis Evangelou
Wasn’t German, and I couldn’t understand anything, right? And there was furious, I guess, my father, why didn’t you stay in German, I haven’t been able to understand all this, right. And I was feeling that there was a huge part of the work that I was making, that everything was like, Greek centric, and around me, but I wanted to get out of this Greek bubble, like before respect to the Greek bubble, I wanted to get out of it. But I couldn’t find any other way. You know, there was no internet. There was I lived in Athens, and I lived in a very small city in a very small town. And we didn’t really have anything, we didn’t have any cinema or theatre or whatever, where you could really expose yourself something different in terms of culture. So it was a I think it was kind of early reaction. It was I need to expand myself. And now, when it comes to my kids, I see that I kind of am I’m not probably too much proactive when it comes to language learning. So, my kids are exposed to languages from a very early age, meaning that they start with Greek for me only Greek for me, only French from the mother, only German in the environment. And so school and so on, and when it comes to watching films, or whatever on their tablets and so on, everything is in English. So starting with passive learning and so on. And, yeah, multilingual brain is something different to become multilingual at a later age. So we all can speak multiple languages, but we became much more most of us became multilingual at a later age. So we missed a few things. Our neuroplasticity is a bit different. You know, we’re there, but those kids, not my kids. The kids of this generation will probably be there. Right. So I, I don’t know, the work need is ideal to this point. I feel like I needed to expand, like, otherwise. I was feeling like I was missing too much.
Andrej Zito
Right? Yeah. I laughed a little bit because to meet like, I imagined it like, like a six year old Yanis comes to his parents and he’s like, Hey, Mom, that I want to expand myself.
Andrej Zito
But actually, yeah, like, do you think you were seen as by your peers as a sort of a nerd, because you mentioned really like to study a lot, which is unusual for people. Yeah,
Yannis Evangelou
I wasn’t there. I was, like, I was like the school nerd. Absolutely. And this is why, when people met me later on to university, people who didn’t know, before, they were like, This guy’s nose is a complete failure. He doesn’t really do anything. Tell me what just football unsalted fails every single test, which was true, it was like, five years of not being able to do much. I was refusing to move on with my life was like, huge. And, yeah, that. I mean, if I feel like, if you haven’t done something in your life, but you still feel like you can do it, like even even if you’re 26, or whatever, we feel like, Oh, you’re 26 and you haven’t finished your studies in whatever your first degree? It doesn’t matter. It’s like, you can. When I started, when I switched from translation to programming. I was 34. I was 34 with a newborn child. So And was it challenging? Oh, yes. It was challenging, of course, because you feel like you’re starting everything from scratch. Everybody around to around you says something like, come on. You’ve said before, like, You’re too old. What are you trying to do? I mean, you’re not 20 anymore. Like, you will not make it. Right. But there’s no age restriction. When it comes to, when it comes to being motivated enough to start something to learn from, and there’s no age restriction when it comes to learning. My issue with a lot of people, I have to say, is that there is this, I feel like they’re never motivated. Or they feel like you know, I have my degree, tentacle, I can live the rest of my life like that I can have the same job and so on. I wish I have the same job. Like, I don’t really need or want to learn anything more than that. For me, okay, maybe the worst, failure is too much. But these are personal failures, like not being motivated at all, to move on with your life,
Andrej Zito
did you? Did you ever think about why people, let’s say choose this, I don’t know, if I should call it the comfortable path, or they think it’s comfortable. Like why they don’t want to expand themselves.
Yannis Evangelou
I think that for every single one of us, it’s very hard to get out of our comfort zones, like, for me to turn my camera around and have a podcast with you. It is hard. And it might be hard for you to to interview other people, like, you know, he said, I’m a bit stressed, but they will make it even if you’ve done it 200 times, right, you might still be stressed. And you get out of your comfort zone in order to do that. But there’s no other way. So I think that it has to do with social examples, or the examples of our social circle around us, we see a lot of people in their comfort zones, kind of enjoy a relatively happy and stable life and so on. We feel like, well, you know, it can still work. Why should they take that many risks? Right? I think if a company moved from Slovakia, Slovakia to Canada, right? You could have stayed in Slovakia, or you could have made move to a smaller village in Slovakia.
Andrej Zito
I can start my own village of one person.
Yannis Evangelou
And, yeah, it’s very easy to reproduce the same thing. It’s very easy to say it is too late to start with. It is too late to do anything. But I feel like the people who were saying when they were 25 there were saying, Oh, it’s too late to do that. You know, they still say certified in 45 minutes on it becomes a pattern, which is very hard to break this kind of but it is hard to break into this. I mean, it is hard for all of us. It’s like being disciplined. Like all three of us. Maybe we say You know what, I want to learn Ukrainian at the moment, right? And we’re starting and three days later we’re like, oh, now I have to do something else. I have to do something else. Two months later. We haven’t moved at all. So maybe motivation is there. But discipline you’re motivated then. Yeah, exactly. discipline discipline is something else. So motivation might still be there. But discipline is absolutely necessary. It’s not only about motivation.
Andrej Zito
Okay, good. I think we got you talking Jani finally. So I think I have to switch between you because otherwise it’s going to be like coming to be one person. And the other one. But no, I’m, I like this topic. So what do you think of a silly about this thing like why some people choose to stay in their lane, and they maybe don’t even have the motivation to to expand.
Vassilis Korkas
I think if you asked me, for someone who is younger, it’s probably an issue that they don’t feel like they don’t want to look outside their bubble. But the older you get, the easier it is to feel like you have enough and you’re happy where you are. And you don’t want to take the risk. Or you don’t want to apply your time on some activity that is completely outside your comfort zone, when you feel perfectly happy with what you have, with your job with your family, with your living arrangements wherever you live. And that’s all good. So it is always kind of trade off, whenever we try to get something new going. You want to know, in advance as much as possible, especially the older you are, that there is going to be a return on whatever you do that it’s going to be positive impacts on your everyday life, that it’s going to help you or your family or somebody else was close to you. And you need to take that all into account. And if there is an opposite effect, meaning that if you start something new, and you start having less time for yourself to rest for your family to see you, for socialising, for having a work life balance, whatever that is, then it’s becoming more challenging to you start thinking more, maybe I don’t have the capacity, maybe I don’t want to find the capacity, even if that is there. So you sit back. But as young sets in, let’s say in your 20s, or your 30s. When you’re when especially in your 20s, I would say where when you still don’t really know what you’re gonna do with your life. And no matter what you think you know about your life in your 20s, I can tell you right now, it’s got nothing to do with how your life is gonna be later. I mean, this, we started this company, I was 41, I had no reason to take any of that risk, when I left the university at that age. And still I thought, well, whatever I’m doing here is not good enough for me anymore. There are things here that I don’t like anymore. And I don’t want to be dealing with them on an everyday basis. So I’m just gonna leave, we didn’t even have Lex go back then, as an option. We hadn’t even we had started talking about an idea. But you know, that was much before. Any case, if you’re in your 20s, let’s say and you’re still trying to figure out what you’re going to be doing in the next 10 years. Well, in your 20s, you rarely think about be doing in the next 10 years, you might be lucky to sing what you will be doing in the next year or two. Let’s say you start with the next two years, the more options you have for yourself, the better it’s going to be, you know, you have no idea where that little white ball is gonna settle on the wheel of fortune, you have no idea. So you might as well have more options where that little ball can settle. Even if you study something that might not make sense right now, it might make sense five years from now, 10 years from now, you might think that you’re studying translation, for example, and some of the topics that you’re dealing with are so incredibly boring, that you want to kill yourself. Don’t do it. Nothing is waste. Don’t kill yourself. And nothing is wasted seriously. I mean, I’ve been things that I sort of developed as an expertise later on in my academic life. As a student, I thought, why are we even doing this? This has nothing to do with what is current? What how the industry is working, let’s say or whatever I thought the industry so it was current. But it turned out that you know, some things are fundamentals. They won’t go away, no matter on how much the industry advances no matter what new technologies come forward, etc, even now, let’s say with empty with GPT, with NLP with all the most cutting edge advances in technology, the language skills are still gonna be there. The fact that you need to be able to rationalise and explain to somebody who is a non technical person, how does this work? And why can it help me? You cannot do that only as an engineer, you need to be someone who understands language. So, you know, there’s always this gap between different specialisations, let’s say, and people think that, oh, that side is always right. And that side is always wrong. That is never the case. There are always complementary skills that you can bring to the tape, no matter how late in your life or your career. And obviously, if you don’t have the willingness to learn, that is not something that’s anyone can really imbue than you. So nobody can just crack your skull open and feed this into you. You either have that or you don’t. So that’s fine. If you don’t, that’s fine. But if you do, don’t let it go bad, make use of it.
Andrej Zito
I actually have a question about that. So I think you mentioned that you either have it or you don’t. I would like to think that you can maybe cultivate it, especially from the early age. And that was actually the question that I had for you. Because young you were talking about, like how you teach, teach your kids, you know, to live in a multi language world to speak languages from an early age. So how would you how or how do you try to raise your kids so that they are willing to learn? And like, in a way you don’t kill their curiosity? Because I think that children in general, like most, most of them, are all them? I don’t know, I don’t have kids. They’re curious. But then, you know, like, as you get older, your curiosity kind of like dice off with many people. So how do you make sure that I don’t know like, once your kids are adults, they will still want to pursue new things, try new things be open to risks,
Yannis Evangelou
right? First of all, whatever I’m doing as a parent, might be totally wrong, right. But this is just my approach based on my readings and my influences and so on. First of all, I have chosen to treat them as, as if they were adults, from a very early age, meaning that, for example, before our podcast, I described what two of them concerns and cannot fully understand, yes. I describe what we’re doing, why we’re doing it, who you are, trying to tell them who you are, and so on. So that they build their own impression of what we’re doing. Once this is, once it goes live, we will watch it together. And I will explain some things we will discuss and so on. Right?
Andrej Zito
If one were to like that, you’re gonna be on YouTube.
Yannis Evangelou
Right, they want to be YouTube. Creator. So what I tried to do I try to employ a personalised approach, which is that kid, one is still the brother of kids, two, and the brother of kids three, but all of them are totally different. You know, they might like total different things. There are some things that they like they have in common and so on. But everything else, yeah, it’s different three different worlds, two different clusters, if you want, there might be some intersections. But otherwise, there are three different clusters. So I follow them, and they jump in, whenever I see that there’s something that I could contribute. Otherwise, I just, you know, let them grow. And they asked me so we have developed a relationship of trust. They asked me so what do you think about that? I do the same when there’s something with regards to the electrical for example, oh, you know what, we have a new client who asked for this and that now you’re nine years old, you can tell me what would you do? And he says, No, you’re you know what, sometimes they told me things because I had never thought of it was like oh of us like so i i Don’t block them. I don’t have high expectations from Last school, I tried to provide them without as many skills as possible. And skills. When I say skills, it’s not only about computers and language learning or books, it also has to do with, you know, they’re being wired to the nature, like, also survival skills, a cooking, cooking, they learn how to cook together with that, and so on. Why, because I feel that those, even if they, even if they choose to leave them in their own cluster later on and become too introvert or whatever, they should be able to, to survive using their own tools and using what I’ve tried to provide them, and what they have acquired on their own. They don’t need to show me that they got what? Well, in German, it’s like, one is excellent. And six is like fail. At school, it’s the other way around, you bring an average of three, it’s fine. It’s fine, because I see that people can afford it, it’s fine. I don’t want I don’t expect you to bring one in every single class, I don’t care. And they won’t care later on. If I see lack of motivation, or lack of focus, or whatever, I will try to understand why there is this kind of luck. So I, quite frankly, I develop myself, together with them. So I try not to be in the distance. I’m not trying to come from a ball and like, impose my beliefs. Because I’m biassed, you know, I’m 43, I believe in totally different environments, and so on, I have different experiences. So I have a bias, which can be also negative for them. It’s not always positive. From therefore, I’m, I still have like a safe, effective distance always with them, or the sake of distance with regards to work to guiding them. And so with regards the skills, I think, because as I said before, well, either you have this, or you don’t I mean, you can always cultivate something, but it’s either you have it or not. What I noticed, just in our industry, for example, but this applies to all industries. I think we have, let’s say a lot of engineers with no language skills, a lot of language translators, and so on with no tech skills. And then if we try to find some product managers, or somebody else who will connect those two, which is fine, right. But I think without most of the engineers do not try to reach out like to the other side, don’t try to acquire any kind of language skills, don’t try to dive a bit deeper into linguistics, like learn the basics, like, you know, you’ve been an engineer in the language industry for 20 years, try to learn a second language type learn the basics of linguistics, just something you know, you don’t need to learn everything. If you are a translator, you’ve been a translator for 20 years, try to learn the basics of you know how API’s work, at least some things you don’t need to become an engineer. Nobody asks you to become an engineer. But you need to understand this, because this has to do with your daily workflow, you are the one who kind of passes information to the product manager, who will have to translate that information to the engineers. But you know what, you can do a bit more than that. And you engineer you can also do a bit more than that, to understand what they are telling you. You cannot just expect the doc, you know, bullet that you follow. So I feel see this up. And another way both sides are resilient to learn. I think that they’re more resilient. I disagree. But I should just go. I don’t know, Andre about your take on that.
Andrej Zito
I don’t want to go into my text because it’s already one one. It’s already one hour in our interview. And we still didn’t even get to talking about when where did the idea for Alexa come from? And the reason why I’m saying this is not that I’m not enjoying this conversation. I really am but I did an interview recently with another founder. And we spent the whole two hours just talking about these topics. And we didn’t even get to talk about his company which which should be in a way like part of the interview. But anyway, I’m just I’m just I’m just letting go. But But yes, I also want to hear from from my CV. I assume you have a family. You have kids I’m or
Vassilis Korkas
I’m married. No kids. Oh, no kids? No, I have a sister two years younger, who lives in England? And my parents back in Greece.
Andrej Zito
Right? Okay, so I will not ask about why not kids yet. But maybe since you used to be a professor, like, how can the professor, let’s say, instil this curiosity or this willingness to learn to their students? How do you keep them engaged? Because I don’t know. Like, I’m pretty sure that when you were a student, there were teachers who were like, super boring, and there were teachers who were maybe excited and can maybe they give you a glimpse into how the real life could look like. So how did you do it as a professor,
Vassilis Korkas
my generic approach was always to have a problem solving elements. Indeed, because that’s engaging somebody at the same level, like everyone who’s in the group, they feel equally involved. And they don’t feel that there is anything separating the more feeling them, I don’t know, second class or anything, it’s the same problem for everyone. Everyone has the same kind of information to begin with. And they all know that they could potentially fail in addressing the challenge in finding a solution to the problem. And that’s not bad, that’s not a bad thing, it can very easily happen. The idea of being in an academic environment is that it’s a safe place, right? Even if you fail an assignment, it doesn’t mean that a customer, you’re going to lose your customer, there’s nothing to lose, there’s always something to learn, but there’s nothing to lose. So by presenting them gradually, more complicated problems, to try to solve, whether that had to do with, I don’t know, difficult terminology in technical texts that they had to translate or trying to figure out how katsu was working, when they had only looked at a couple of things before, nothing more, nothing in more depth, or how to work together in groups to create a company presentation. for group projects, which is something that, let’s say is not particularly common in subjects like translation, it’s more like personal study, rather than group study. My experience is that all of that kept everyone more alert, and more willing to cooperate as well, they started understanding that a lot of the resources that they have, are out there, they are not just in their heads, but there are things they can reach out to other people to find. They can improve their research skills, whether online or elsewhere. I mean, at the time, when we started online, was nascent, there was practically nothing on the internet, and even then whatever was there, looked like, you know, Paradise. I remember, when I was studying, the dictionaries, we will find online, would come through Alta Vista, with some very basic links. And we will go on, I don’t know, some glossary that someone had put together, because they, they had the background, let’s say in specific fields, and we would see that glossary online with definitions and thinking, wow, okay, this is the future. And now you don’t you think of dictionaries, and you’re thinking, what’s the point? Why would I need a dictionary now as a translator? Well, there are situations where you would need and you would also need to know how to use it. And there might be people right now for donati’s addiction. And you’re thinking you just flick through a page or you find it online and it’s there for you. But decoding the information that’s in the dictionary can make or break, a choice that you will make in a translation, depending on the context, depending on the subject fields. And it’s the same for figuring out terminology problems, for example. So these are, let’s say basic skills that you would expect anyone to have, but they don’t necessarily get the chance to develop them unless they get involved in problem solving situation. Do they have a problem solve where those skills are required? Yes, then they will need to develop those skills in order to solve the problem. If they don’t And then it will be an unknown until later on in their career, they might have to deal with it. So, yeah, I mean, there are certain, let’s say subject fields within the field of translation studies where I thought problem solving can apply a lot more easily. And that challenge is there basically, for everyone teaching translation at any level. And any particular, let’s say, subject, people on teaching, let’s say the more theoretical modules like translation theory and things like that, that was always the biggest problem for them, trying to find something that will connect the theory with the practice. And that’s why a lot of scenes, I’m sure Yanis can agree with that. So that translation theory was one of the most boring subjects when studying translation or that level, but that was because it was very difficult to find a practical interface, something that will make them think of how to solve the problem. Practical translation classes have a lot more opportunity for that. So there might have been more engaging by comparison,
Andrej Zito
do you? Do you agree with that Yanni? Or were you the only student in centuries who was like yeah, translation to you?
Yannis Evangelou
Know, I, I, I was not. Well, yeah, I was not. I was not supposed to be a nerd anymore when I went into the UK, but yeah, you
Andrej Zito
were cooking books are right. Yeah.
Yannis Evangelou
I was still a nerd. And this is why after watching this, my dissertation was in translation theory, like in neologisms in philosophy. And my PhD, which I dropped two years later, because I had a full time job at the same time was about the same thing. Like comparing neologisms in philosophy. Anyway. But it’s true that nobody thought most of the people wanted to have a master’s degree in translation in order to get a job as AP. And it was being well back then. So the rates were high, there was no machine translation opportunity. Nobody was everybody was making fun of systems. Nobody was even talking about machine translation. Everybody was saying, system, you know, it was? Yeah, it’s the brand name had become a synonym of the technology. And All right, we’re gonna call and it might be a Pepsi or something, and see the coke. And what surprised me though, was that most of them didn’t become translators. So they went back, they were going back in, just one year, they were going back and they were not becoming translators, they were keeping this master’s degree in order to do something else, to show this as an extra certificate of skill. But when they were facing, as soon as they were facing the reality of being a translator, they were just holding back and was like, Okay, I’ll, I’ll probably do something else. Because even if the rates were high, being a junior translator meant that you still have to work like 10 hours a day back then. And at the end of the day, you would get like 1000, you’re a brother of the end of the month, like working with a translation agency, right. And if you want to become a freelancer and impose your own rates, and so on, you would need to have clients, which means that you needed experience you did the network, and so on. So take some time, who has the patience to do all that? Not many people, right. And it’s not only about patients, sometimes it’s also about priorities, you know, some people need to make money as up, they have different needs, and just cannot wait five years now to know to build a network in translation. Now, obviously, while people think that it’s even worse, I would not say that it’s even worse, a lot of I’ve been translated for I was transported forums a decade. And I was also thinking back then that a lot of the translators were spoiled, meaning that they, you know, it was like, I’m an expert in game localization. If we were going to have translated and on to interfaces in different games or something like that. I want 15 cents for translation from English into Greek or Slovakia or whatever. Why’d you know? There’s too much? I mean, it also depends on it also depends on the content, what kind of expertise do you bring, because if it only about translating the words, New, Open the window, and so on, and you don’t really bring any expertise, you know, you are you can be replaced by your it’s not only about technology. And there are people like that there are still people like that who think like, you know, I’ve been in the industry for many years. So my rates will still be high, will always be high. But it’s not just about rate. So the rates are always related to the content, they’re always related to what we get. So back then it was something totally different, you know, you needed a future, you need a bunch of dictionaries next week, you barely had an internet connection to send the files, to upload them to an FTP to take your one hour Phoronix makes, right. So things were different. And now we have continuous localization, we have, we have empty, obviously, which is not something to laugh about. So the quality of empty has not now has nothing to do with the contents of empty 15 years ago, or 10 years ago, or even 10 years ago. So we have to adapt, we cannot always have the same. We can always make you cannot always maintain the same claims about us or about the world around us, the world changes much faster, you know, we cannot be static. I would also love to be paid 100 euros an hour to make just an HDFS load some CSS main HTML page, right? That’s not the case anymore. Okay, I have a clip added. It’s fine. About button today, things were changing. I remember. My first 15 years ago, I started joining the UX group usability group I was interested in because I didn’t know how to code. But we started building some WordPress websites. And people were like, Oh, now there will be no need for web designers was like web designers back then, like, nobody was talking about web developers, right? And nobody will need that web designers anymore. Everybody will switch to WordPress. Yes, WordPress, got a bunch of traction. But we also had a new market. So there were word plugins, WordPress themes, like ThemeForest, a lot of people made money on simple. So there are a lot of new opportunities. And it’s insane for every advancement in technology, when we talk about empty charts, GPT, WordPress, whatever. It also offers opportunities. It’s not only about what we will be missing from our previous steak, our rates might change our technology, our tech stocks might change, a lot of things might change. But it cannot only be negative, we shouldn’t only see the negative aspects. And I think that we’re often too selfish. And we were often too selfish. And we think that so selfish and biassed. And we think that when something affects our rates or our comfort zone, we think that it’s negative, we don’t we refuse to see the positive aspects or prospects of what may come from this opportunity.
Andrej Zito
Yeah, I think it could be very much related to what we were talking about before, like people’s comfort zones, right? Because like this new advancements, basically, I don’t know, crash, your idea of your comfort zone and like what has been working for you. So maybe that’s why people are hesitant. Anyway, speaking of advancements and innovation, let me quote something the conceptual seeds of lexica were sown soon, I don’t even know how to say the words. So sown back in 2013, when Yanis realised that there is a gap in the localization industry, can you give us quickly a context of what you were doing at the time like professionally and what what was the gap? How did you realise the gap?
Yannis Evangelou
So I was trying to make it as quick as possible. Get my first job as a QA specialist in LSP. And the UK back then, and they were what they were doing. They were localising all Bentley manuals. The cars or the elevators, or Bentley manuals into 13 languages. And a friend of mine who used to be a student over Phyllis, work there. So I asked her like, so how do we, how do you do QA? There is no more, there’s no QA, we just, you know, we translate, we send things out, and we get them back and we print them. And I said, Okay, I have an idea of my talk with the boss, and said, I don’t think she will accept you, but I will give it a try. Anyway, he made it and I had half an hour with her. And the boss said, What did I say? I said, You know what, there are some QA tools. And back then it was very basic, there was QA distiller, like first version or whatever, okay, distiller errors, py, and expunge those three, I’m not wrong. And I said, you know, we cannot make a part of the process. So what you’re getting back, at least, at least, we can have generic quality control to make sure that what you get back from the Chinese or the Hindi translators doesn’t really include any critical errors, at least let’s avoid the critical errors. And she said, Look, I don’t really have any position open or whatever, but I will give you one month to prove me that this thing works. Right. So it’s got a long story short, second day in the company. And there is a party, there is a morning party, everybody’s happy because they just tend to the printers all put in manuals. In fact, it was not just certain manuals, it was three manuals in 13 languages. It was 39. Manual, right? And it was something which would cost like 1.3 million pounds, right? And where was so I was still a nerd. Right? So you party, but where is what were those copies? Like? Have you sent to the printer? And it’s in the bin outside? Okay, I getting bailed out, they get into the bin, I get the copies out and say, sorry, sorry for ruining the party. But since two out of the three manual, the brand name on the front cover is wrong. What so they started cursing, obviously, and firing everybody. Anyway, anyway. And yeah, so they had to pull everything back. She called me to your office. She said, Okay, now you have a contract. I mean, I had I had done manual QA. Still nothing automatic, right. But then I convinced her and we bought a couple of QA distiller licences. And so we automated the parts of the process for the next printings and so on. Oh, then I started. Working as a translator, after I left, I went back to Greece, a few months later, I started working as a translator, I noticed that nobody was doing any QA at that translation agency, which was supposed to be like the number one in Greece. And then I moved to SDL. In 2009. As a contractor, I was mostly doing reviews and evaluations, what it was like full time. And again, nobody was really doing any QA. So okay, let’s talk with their, one of their big clients. Like it wasn’t team lead for rather big clients. And shouldn’t we use? Is there any QA mechanism on your platform? No. All right. Do you think you can use that they had an online they have built their own online platform? Do you think you can use that? Well, no, because this desktop tool, so we cannot really use expensive QA distiller and so on. We could use them with a cart, but we don’t want people to download our data, keep them and so on, you know, we don’t want them to be able to download CMX and reuse them. Okay, there’s nothing like online QA. There. There was nothing back in 2006. There was nothing in 2009. But now it’s 2013. It’s late 2013. Somebody should start something. So I invited four friends of mine to developer some tooling, which could start a company which could build something like that. I don’t know how to build it. I’m not the developer. But you are the developers. We can give you all specifications, and maybe we can build the first version like, you know, better or something like an MVP, MVP. And they were interested but all of them were full time professionals and And they didn’t have any one.
Yannis Evangelou
Yes. So we, we dropped the idea quite easily because there was no company and so on. But I also decided to drop translation and become a developer. Because I just didn’t want to be there anymore. Like, I didn’t want to be working in something like a factory. It was like a factory. Like being in a factory. I had, you know, custom deadlines. I was feeling almost burned out. And then my mother said, oh, you know what? Your brother in law. So Demetrius was married my sister. He’s really good at computers. He’s really good at computer. She didn’t know he was an engineer. So he is good at computers. Anyway, you can talk with him. Anytime I talked with him. I said, You know what, I had this idea. What do you think? Oh, yeah, we could try it. And I was always talking with my significant on and she was about to leave academia. And I said, You know what, I’ve been checking some. A lot of people get fundings. So I started product Product Management for for a couple of months. And there are multiple ways to do it. And especially in Berlin, back then it was like everybody was getting funded. And we applied to several institutions. And we ended up in Rome, where translated, was also part of, to a halt, which was offering fundings and translated, so the people from translated, saw a presentation we will discuss, and they were already planning to build something custom. So they were like, well, you know what, you already have that demo, if you can execute and have this in six months on makeup. Let’s try to do it. So we got the funding from them. And we did the integration with makeup. We started with four languages in June 20, Italian
Vassilis Korkas
and German, right
Yannis Evangelou
was in June 2016. And the lexical integration is still active, but now we support 160 locales a couple different, the scale is different. And so this is how it started. It was more like that there is a gap that nobody wants to feel or tries to feel for any reason it. So 10 years ago, it was generic. 20 years ago, it was generic, like, Why does not? Why does nobody take care of it? And we decided back then to take some risky decisions. For example, we said, we will only be cloud based, which will be API. So imagine that was December 2016. In fact, so things were not like now things were totally different. People were using the extension, very couple QA they were using multiple truthers and MOQ for translation and so on. So Memsource is still an early phase SDM, Memsource, and so on. They were still like, early cloud based. Try those copies if you want, in a way, like trying to move crowds, close to the cloud. And so many years later, that’s the question. Still, nobody really deals with quality control. So a lot of people talk about quality. A lot of people talk about quality, about LTA about quality assessment. And easy. It’s a bit easier to talk about quality assessment because you don’t really need to do a lot of tech work, meaning that you can build a platform to support quality assessment. You can have a lot of discussions on panel about quality assessment.
Yannis Evangelou
still like something that the lawsuit will try to hide under But like, we all know that it’s challenging to invest a lot of time in building only one module, what I mean is like, cloud based tools have now become, like dinosaurs. They’re huge. So let’s say Memsource started with five features, and now has become phrase in the cab 100 features, right? If they need to add a reliable QA module for all the languages that support, they need to spend at least two months, years, at least, to my ears. But more than that, but anyway, so yeah, it takes a lot of time. Well, we only deal with that module. So we decided to take the risk early on, not to not to spend any efforts on providing services, for example, like, we will not start with QA and become an LSP, or whatever, we will not start with QA and then expand to become a mini TMS or something. We don’t want to compete against the MSS LSPs, or whatever. We just want to build that QA module and make it as reliable as possible. And as customizable as possible. Why am I saying customizable? Because even when you use an empty engine, you might use, you might say, oh, dip a little bit. And then you use the title for a field where it’s course only 33%? Well, precision is really low. And something else that scores 95%. Right? It’s not an off the shelf solution. You still have I mean, for me, it’s a bit different. But when it comes to tech, in general, it needs to be customizable. And so we try to make everything API first modular, local, specific, customizable, and we try to change something in the workflow, which was in traditional QA, you finish your translation, you click a Like a QA button, there is a QA report, and you have to go back and forth all the time on baby. Well, best case scenario, there is a side panel and you just go back and forth in your editor into QA. But what do you do there? You think, slow, you don’t think fast? Because you rapidly read descriptions like segment one possible error in number one, it should be spent time reading and they said, Oh, it was a false positive and move to the next one. Okay, again, again, a false positive, and you get frustrated and you never want to do any QA again, right? That’s one thing. And the other thing has to do with the workflow like I don’t know if you use Grammarly, when we, when we did our pitch, eight years, almost eight years ago, we presented ourselves ourselves as grammerly for translation. Just because we were introducing this live QA thing like archetype. It’s not that we were checking for syntax or something was mostly about the engine running in real time, meaning that it will it will also act as a time saver first. And second, the translator doesn’t rely on any reports or whatever anymore as they typed the see the errors they fix them. And then they don’t have any excuse to say what up I missed it. Oh, it was just in another report. It was too too long to read. They had a deadline or whatever now. It tells you type. So would you use Grammarly if you had to save your email or document download it upload it to the Grammarly server and not like that. So we try to make it as intuitive as possible. And of course, it took some time until we get to the market, meaning that we started with our makeup integration. But how do you sign the next client then you need other integrations? That’s one thing. But bank of everybody pretty much everybody was using traditional MOQ we didn’t want to do in integrations with with desktop tools.
Vassilis Korkas
Because not only we didn’t want to it’s not easy to make them. What
Yannis Evangelou
are the steps in the workflow
Vassilis Korkas
but you also have to have a partnership with the platform. And you know desktop to online or online to desktop doesn’t work very easily. You always have To switch to a different environment in order for it to work. So we want to
Yannis Evangelou
nobody wants to change. So everybody wants everything, in fact consolidated into one environment, if they, you know, they want, they’re empty. They’re the project manager and their QA, they’re everything, can it all be in one tool, then we will use it. Even if you have the best tool in the world. Nobody wants to do that extra click opens up extra tab, you know, extra login, and so on. Nobody. So it took us a while it took us, I have to say it took us a couple of years to first to add more locales, because you cannot really be a QA tool with only four locales, you need to be able to cover at least tier one, at least one thing, and the other thing has to do with with integrations, like Do you do any integrations? If yes, with whom? And do people actually look for extra integrations? Or do they expect everything from their customers? And, quite frankly, I’ve noticed that they expect everything from a cartoon, everybody complains about their cuts tool of their choice, like, you know, they do long RFPs they pay a lot. They end up with tool number one, and then they complain about what it lacks. And you cannot have it all, obviously, but it’s as if in. So, back then we had truthers in Malmo, two people tried to build a clone of those, or the cloud. But the thing is that when you try to add more features later on, and you are not proactive, meaning that you have not really cope about adding more features, like you try to go to the market, you say, Okay, I’ll build something similar, but on the cloud. Now, I have to add more features, more steps in the workflow and so on. And this gets out of control. Meaning that also seeing some of your live experiments on your channel, when you were testing various TMS. There are popular TMS environments, in which even if you have 20 years of experience, you might be spending 10 minutes just to start the project just to find where that button is where you need to start a project or open a project. So it has gotten out of control, meaning that imagine being a translator, and having to learn all those tools from scratch, because you have to work with a lot of tools at the same time. It becomes frustrating. But when it comes to QA, unfortunately, not much effort is put. So it’s always like, I don’t know why it’s not in the top five priorities. But it seems that it seems that it doesn’t seem the second top 10 Probably, but definitely not in their top five priorities. And they all keep avoiding discussing LK. They do a lot about quality assessment, because it’s easy to say, You know what, we’ll have some will at the forum, we will you will be able to add comments and evaluate and so on. Right. But may I have a bit more automation? This is where it gets a bit tricky.
Andrej Zito
I’d like to go back a little bit
Andrej Zito
going steel back to like when you brought this idea to to a series I think you mentioned that was really wanting to leave academia. Why did you want to leave and what were you considering as an option before you came with the idea?
Vassilis Korkas
Well, this is an interesting story. The the main reason that back then I felt it’s time for me to go is when I realised that academia I At least in the UK, and the way that I was personally experiencing it had started losing. Its, let’s say, lustre. So this appear the loose there is like this veneer this image, it’s, it’s something like an ideal that you have about something. But then you realise that has worn off, and it’s not there anymore. And what sort of tipped it for me? And made me realise, okay, things are not going in a direction that I like, was either an internal meeting, I don’t even remember when, when somebody started talking about students as customers. And I thought, okay, are we really using this kind of terminology now internally. And mind you, this was not by any of my colleagues at the academic department, it was one of the administrative parts of the universe. Nevertheless, that gave me a very negative impression. And then I started realising that there are things happening in the backgrounds, from the various departments, which I have no exposure to. But inevitably, at some point, they start spilling over. So you start getting that kind of exposure as well. And this creates a very negative feeling for me. The fact that it wasn’t a school anymore, it was a business. I was there because I was, I really liked teaching. And I always saw it as a school as that kind of environment. So when that veneer disappeared, I thought, Okay, well, I think it’s time for me to start thinking about leaving, even though I didn’t have any idea about what I will do next. For a year or two, I switched to part time teaching. So I was I didn’t have my permanent post anymore. After I gave it up. And during that time, I was just going at the university four to three times a week to do a few classes, and then I would go, I wouldn’t have anything to do with what was happening in the backgrounds. And I dabbled in photography for a couple of years. Part time, that was one of my hobbies, back then, which I thought, who knows, maybe I’ll try to develop that professionally. Obviously, that didn’t work. Because, you know, if you don’t really commit yourself to it, nothing is going to happen. You really have to be there. 100%. So it was just a hobby, a full time hobby, let’s say, but it wasn’t really a joke. And then that was the around the time when, when Young’s sort of shared that idea with me. I thought, Okay, I think there is some potential here, because this is not something, you know, that in my experience, both as a teacher and practitioner, limited experience. So as a practitioner, I could see that there is a potential there to make a difference. Had nothing to do with running a business or understanding technology, how that worked in the, under the hood. So my angle was completely from the linguistics side. So if we were to say, develop QA checks for language, what checks should we be worried about first? How do we put them together semantically, how do we combine the languages together? How do they make sense in a system that a user will be able to understand? So that was my personal involvement at first. But then after we started the company, then, you know, things evolved in a slightly different way to I mean, I had to learn things that I had no clue about, like running the company finances, or, you know, writing content for website for documentation for everything. But, you know, I was very much motivated by all that because I knew there was a greater goal in the background. And it was always something like every day would bring some Thank you. So we were learning something new every day. And that was part that goes back to what we were saying earlier about the curiosity that I was curious enough, even at that age and that stage of my life to say, okay, yeah, here we go, let’s do this. And I’m not saying that it was easy all the time, right? He was far from it, it’s exactly the opposite. It’s like the hardest thing I’ve ever done. In my life, even moving country twice wasn’t as hard as this. But there are rewards that are always implied in a way, you don’t talk about making a huge salary as a reward. Because there isn’t such a thing. For a small startup. There are other rewards which make you feel like there is a sense of fulfilment, there is something that you are achieving. Even if you don’t have 100 customers, you still know that what you’re producing is good. And as you can make a difference.
Yannis Evangelou
On top of just a couple of notes, one is that, at the end of the day, Andre, it’s not about who has the ideas, you know, because it doesn’t really matter. It’s about it. Because all three of us, it could come back the other way around, like bacillus, or Demetrius having the idea doesn’t really matter. For me, it just started because I was facing that I had been facing that problem for many years and didn’t see why I couldn’t see why anyone couldn’t solve it. And it was mostly about all three of us agreeing to leave our jobs, which were, you know, what I mean, we were making quite good money with those jobs. Taking the risks, at the moment that it, it could be challenging. Like, I have the newborn Demetrius, for example. My sister was also pregnant, and he would have kids, and then we had another three kids during electrical, so it was like, a lot of challenging times. So it’s mostly about people coming together, and not just having ideas, but also executing and insisting like being persistent, at the end of the day, you know, what, I know that there is this need. It is getting clearer and clearer, we will get there. And I said the first two to three years were much more challenging, obviously, until we build a system which would support quite a few locales, and that they would continue to build the first network of leads, who would then convert? The thing is that the our case was then confirmed so. So especially after 2018, when we noticed that even people who were kind of ignoring us back in the beginning, like, you know, what, we have travelled, we have an MOQ or even while you’re under the cloud, but you know, we don’t need anything extra. They were getting back to us. And they were saying, You know what, we need to try this because we have issues with QA. Now we understand what you meant, like two years ago, that there could be potential issues. So linking this back to what we were saying about motivation, curiosity, and so on. It’s also about persistence, like, not just insisting on something which just because of your ego, but insisting and being persistent and something that you see that you’re very clear about, like, I can see it very clearly. I cannot be that far off. Yeah. And so and the other thing, the second note was about, it’s the difference between being an employee and growing your own business. It’s not about money. So you risk a lot. Most of the time you’re frustrated with everything. You cannot spend, you know, you cannot spend any time you cannot take two hours off just to watch a film or whatever because you keep thinking about clients technology and so on. Your mind is always there. But at the same time, you have the opportunity to grow a team came around your company, based on some principles like what we call company culture, based on some principles that you support. And what we have in common facilitators tonight is some common principles, which is not that easy for Greeks, like three Greeks together, finding common principles like common set of principles. It’s not the easiest thing in the world.
Vassilis Korkas
We don’t work in the same office every day.
Yannis Evangelou
And we have also, we also have worked with different characters like Dimitri is, I don’t know, much more introvert in a way, Bacillus, like, it’s like the Coulomb force. I’m much more like the crazy guy. Yeah. I guess I’m just going to try to control concerns. And like, as soon as we’re done with the interview, I will be big calibre. I could have been a bit more. Anyway, it’s not always like that. But
Andrej Zito
I think it’s a I think it’s a good balance to have a ceiling because of a silly seems to be very calm, always has the answer. And you’re more like
Vassilis Korkas
somebody needs to fill the gaps. Andre? Silence.
Andrej Zito
I had the one question. Hopefully, we can answer this quickly. Because maybe it’s not that interesting for you. But can you explain to me why were you looking for capital, like the investment is I’m not sure I’m not sure if at that time, the whole lean movement, you know, lean startups and being lean was already popular or not, I don’t remember it. But explain to me why you needed money. Because especially since you’re three founders, and you have one who’s a developer, like you, you can do a lot just on your own time.
Yannis Evangelou
I may I take this one. Just because you mentioned lien holding was like a buzzword. When, when I stopped translating, because the first of May 2014, nine years ago, and I applied to the startup Institute, it was an American private initiative, which was offering private education with regards to product management, back end development, and so on. It was yeah, it was all about Lean development back then. So the word lean was becoming popular at that time. Well, I don’t know about other countries, but in Berlin, it was so hot back then everybody was talking about lead, then it was Agile and Lean those two, Agile and Lean. Right. And we were bombarded with those two terms. To degree that I, I never used them after that. But one of the things I learned there was that even if you have sold, let’s say you, you’ve had two companies, you’ve made two huge exits, and you have like 10 million in the bank, when you start. So this was their philosophy. When you start once you start your third company, you will look for funding, you will not spend any of your money. And it’s not because you don’t want to spend any of your money, because you want to keep it safe. Because about that, it was because when somebody else gives you the money, you feel a higher degree of responsibility. And I was convinced, to be honest, in the beginning, I felt it was BS, like, what, what does he say? Like, I’ve saved some money, and so, but then seeing how easily I was spending my money that saved money in between, without really having any responsibility towards me towards myself. It was like, Okay, it’s my money, I can do whatever I want that like, you know, in the next few months, I want to be able to do this, this and that. And that was thinking, right, but if that money came from an investor, good, I would have done the same. No, that would have been a totally different story. So that was also a challenge for us. It was something I’ve never done. So if somebody gave us 30,000 What can we do with our currency? How can we so we have to have Good without too, let’s say built an integration in six months. Right? We built it in three months. Oh, so now we’re three months ahead. So it was making us much more responsible in our first startup effort. And I was then convinced that this will be the way to go. It was not. As I said, In the beginning, I wasn’t. But I was gradually convinced that this way you become, you can act more responsibly,
Andrej Zito
was that a mutual agreement between the three of you, it was to seek capital,
Vassilis Korkas
it was, I would didn’t really see any other way of doing it, to be honest, in the beginning. And the same thing happened even a bit later, when there was an offer from our investors to get another piece so that we could use that to grow more quickly. And at that point, we refused. We preferred to grow organically, because we felt that it was that approach was better suited to our characters, let’s say we didn’t feel comfortable giving such a big piece away. And then a basically lose control of how we would run operations. Right. So, you know, it’s, it works in the beginning, as the additional push for us to get things started more quickly. But then later on, when we thought that the same kind of involvement might have a more negative effects, then we kept away,
Yannis Evangelou
and we applied back then we applied for testing. There were many accelerators. So we got two offers. So two out of 13. Most of them, most of the others didn’t reply at all. But this is something that you have to expect when you are when you apply to an accelerator, which has nothing to do with translation localization, some don’t even know the terms like once localization, it was better to apply somewhere where it makes sense. And but as Phil said, we didn’t we have the we could have gone for si de funding, for example, later on. But that would have not been asked. So it has to do with how you’re structured. And we preferred the organic way. And we also felt that the timing was not good. So like, let’s say having a seed funding to build an online API based QA tool back in 2016. Like it was convert 2016, for example, when we were checking that, like, okay, nobody really uses cloud yet that much. Let’s spend that time to build the lexica engine. And if nothing work works in the next 12 to 24 month, then we can consider funding ground, let’s say, but it worked before we consider it. And we get moving organically. After that.
Andrej Zito
You can oh, by the way, it’s already two hours. Are you okay to continue? Or
Yannis Evangelou
if you’re fine?
Andrej Zito
Yeah, I’m fine.
Vassilis Korkas
It depends how much ice cream Yen’s has stopped. I
Yannis Evangelou
think it’s gonna get him like it was like 3300 grammes there and I think you think the package enough?
Andrej Zito
So how much time do we have left before he needs more?
Yannis Evangelou
Leave? There’s no more ice cream? He will.
Andrej Zito
You can choose to ignore this question. I’m curious, how did you distribute the shares? Like what is the share ownership between the three of you and how much is it for the for the investor?
Yannis Evangelou
I think that with regards to investors and advisors, because we also have three advisors based on based on the agreement, this investment, what’s called this investment agreement that was in that’s it. An agreement? Yeah. And that’s an agreement. We’re not allowed to share details. It mostly has Yeah, it mostly has to do with them.
Yannis Evangelou
If you want a mean direct answer, we have tried to be a as democratic as possible, everyone. And since the beginning. And when we, when we felt that something could be a bit more fair, or adjusted to be fair or something, we’ll discuss it like, this is what we do every time. We don’t agree on everything. But we’re not here, or we’re not here to make more money is not the number one goal. So we’re here to build something. It’s what we said last time, it’s about problem solving, right. And since we have a common goal, like to make this thing grow, like to try to solve this problem, we cannot offer a perfect solution to this problem. But try to solve as much as possible of this problem. We understand what each one of us offers, or sacrifices and so on, and we’re all on the same page. So we haven’t, we have never had issues with that. In Berlin, for example, I don’t it’s, let’s say something like the European startup capital. And I remember back then, in 2015, we were said that we were told that a new startup is founded every 20 minutes. Of course, another startup closes every bankrupt every five minutes. That’s another story. And a lot of the people I knew back then started, founded a company with people that they didn’t really know, it was more like, Oh, you’re good in business development. I’m a developer, let’s work together and you do marketing, okay, let’s work together. But they didn’t really share the same principles. They were not on the same page, in general. And everything was, everything ended up as a disaster as a disaster. Although the idea was good, the technology was great, and so on. So it’s not just about the product. It’s also about the people behind the product. And this is how, why, when we built the whole lexica team, it was more like, every time we add one person, that should be a person, we feel confident that they share the same values. Whenever there is a failure, they will not fire back to us or whatever. And this has worked.
Andrej Zito
So speaking about the common principles that you share with our dos
Yannis Evangelou
conference, you mentioned the human overfill if you don’t remember something at the beginning of our conversation about work life balance, right. I don’t know, is there such thing?
Vassilis Korkas
Right when I mentioned it, I left I don’t know if
Yannis Evangelou
I don’t, I think there’s no such thing. But let’s say that there’s approximately such a thing as work life balance. This is different for everyone. And which also means that the daily needs of everyone are different. So a woman who has a, like a female employee with a newborn has different needs different schedules, seeking to work different times during the day, and so on. If you think that she’s valuable, and she can work with you, then you don’t care if she will work nine to five, she might work like two hours in the evening, two hours in the weekend, and some chickens cleat today, as long as there’s no deadline. And we don’t really have deadlines, because we’re not a services company. Now we build our product, who have broader deadlines, meaning let’s finish this in the next three weeks, or whatever, right? But if there’s anyone who has a deadline, it will be the developers most of the time, like if we have to solve to fix a bug or do something else for a client, but otherwise the rest of the people don’t have deadlines. This means that when we tell them when we agree with them that they will work for 30 they will be working for 35 hours a week. No extra time. We don’t count the time. So we trust that they will be spending that time which might not be certified sometimes it’s benefits, but if they managed to do their job in 25 hours, that’s good for them, you know, we don’t need to do any micromanagement. So we hate micromanagement. So it was more about being able to trust people being able to, to experiment also with it and see what did work if Andhra if if I work with with Andre that Well, I don’t know, Andre, I don’t know how he works. But based on our discussions, let’s say, I trust him, or I start to trust him, right? So let’s try this, Andre, let’s work together, we will be having a couple of meetings a week or whatever, whenever you need to, and so on. We’re mostly responsive with our mobile phones all the time. So you can pick that anytime, whenever you need to. But you don’t need to worry about the rest. You have to finish something you have not. It would be great if you could finish this by Friday, or how much time do you think you need? I think I need three days. Okay, let’s take four. And let’s talk again in four days. And if you need anything in between, we’re talking about that this has worked for us. And this is totally different from what we had experienced in translation agencies and so on with a tonne of micromanagement. And deadlines are a reality, you cannot really avoid them in a translation agency. But micromanagement is something that you can avoid. Or you can reduce to the moon to the minimum. So this is one of let’s say, if we summarise this principle, it has to do with trusting that you work with adults, meaning if you work with mature adults, who will not be trying to cheat you. And if they try to cheat you, you know, but for them, because one way or another, you will realise it sooner or later. It’s always like that, in every in every field. This is taught this principle of trust, if you want and good faith. Some people might call it naive. It might be naive, I don’t know. But for us, it is about good faith and trust. This is the main principle. And the other one has to do with transparent muscles will be able to add more obviously, for me, it’s also about transparency, like not trying to find any, any excuses or cover up or make any cover ups and so on. But try to be as transparent as possible. Not everybody needs to know everything about your company. So your junior marketing manager doesn’t need to know everything about the status of your company. But they need to be able to understand what your company does, to a certain extent. So always try to share as much info as possible. Some things always remain confidential, obviously. But you cannot isolate them in their own, let’s say Junior cluster, where they have access to no information at all right? Try to create connections within the company without been distant or whatever. Laughing sorry. Last thing, which is not really easy for anyone, I think when they start the company. When you start hiring people, you often expect them to work as much as you do, or to be committed as much as you do. And even if we don’t admit it, most of the times we’re like, okay, but they didn’t reply. It was like Friday evening, or whatever. Like, we feel like okay, we shouldn’t be like that, but it’s a it’s a continuous struggle, but you should never expect someone who works for you or with you to be as committed as you are. If you are overly committed, that’s fine. People have different degrees of commitment. They also bring different a different value to what you do, you know, you haven’t you have hired them because they can do something that you cannot do. Right. So it’s not only about commitment, but the end of days, you know, like network, shut off services. More principles.
Vassilis Korkas
I can’t really think something else that would be more central to all that. I would probably only add a that any success that the company has had is everyone’s success. And we tried to share that with everyone. There are many different ways to that you can show that you can raise somebody’s salary, even when you don’t think they actually did anything to deserve it. Right. But they can, you can still afford to get a give everyone a bonus, for example, if you got to hit big contract, everyone is contributing to that kind of success in their own way. And the other part, the other thing I was thinking of, while Yang’s was speaking now has to do with the relationship that we have with our external collaborators or contractors, because over over the years, we’ve used dozens of people, in order to be able to build the checks that we have for all these different languages. And our experience, know, when we were on the contractor side, has been generally very negative, let’s say, with most of the agencies that we had to deal with. So we wanted to make sure that we always have a much better relationship with them. So is the job done, you get paid on the same day that you give me your invoice, for example, there’s no reason to wait for 60 days to get paid. You have done the job? So here’s your payment. Do you have any questions about anything that happens during the project, if I can’t help you, somebody else from the team can help you. So always an open channel, you know, you don’t treat them like some external partner who doesn’t need to know anything in the you just let them figure everything out on their own. And that’s has that I don’t know it has created for me, at least every time we have this kind of back and forth with external partners. They always look to work with us again, because of the that experience that they’ve had before. And it’s not, you know, you hear all the time about translators call, I never want to work with that agency ever again. Even if they give me more money, I still don’t want to work for them. Well, that’s one of the things that we really have tried to avoid very hard, and with a lot of people, because we know where they’re coming from, and we know what they should be expecting. So if we can provide it, why not? Right, it’s the right thing to do.
Yannis Evangelou
And we personalise that part of what or our approach to that part, Andre, meaning that we’re not, it’s not that we have like a pool or a database of linguist for Asian and African languages, let’s say. And then we pick the one with the rate, or the dateline, depending on the race or the deadline. And from where we find someone, we would have a call with them, which is often a long call, like to understand their background, like what they do and so on. And imagine being in living in Ghana, or Thailand or whatever, and filling that job. Somebody sends you the project, but they don’t care. Like what we had, in our translation days, we were receiving something common MLD like, Okay, this one you delivered by tomorrow, upon and that’s it. Nobody knew anything about us. We, as I said, we try to develop those relationships. We don’t want those one off relationships. We don’t want someone just to deliver something. And that’s it. What we try to build up, and it’s not surprising that quite a few of them. Even your three, four years later send us like Christmas wishes or whatever, but personalised ones like not just chainmail and they don’t ask for projects and they know that we’re done with glad to tell you that for one more time, but I really enjoyed that. Those three weeks working with you forgot for us. This is a huge world that we you know, we don’t really exchange for having
Andrej Zito
have so many notes here, trying to think about which one to ask. Explain to me so since we’re talking about how you deal with the contract is how does the collaboration actually work? Because also I think when you’re comparing it to the traditional LSP setup, it’s just a different situation, right? Like they handle and work with them on a different level than you do. So from from what I understand is that, I don’t know you reach out to someone and you’re trying to ask them like, Hey, what should we be checking for your language? And then you implement it, and then you’re done? Is that it? Or is there some sort of an ongoing commitment?
Vassilis Korkas
The commitment, initially, for us to develop, let’s say, the checks for a new language could take weeks, it’s not something very simple to do. Then, as a next stage, perhaps, if we, let’s say, decide that we need to invest more time to fix up a dictionary, spell checking dictionary, let’s say in that language, and for low resource locales, a lot of these languages don’t have any, any anything that you can use. So you have to build it from scratch, that takes another few weeks, at least, to build this kind of involvement as well, especially for some of those, let’s say more exotic locales, The linguist who takes part in that projects feels like they are involved in a project that is special for the language. It’s not just the translation, that they’re going to finish tomorrow, and then nobody will remember it. It’s something that will be there. And it’s something that they will be able to use, let’s say if Lex guys integrate, let’s say in a tool that they use somewhere. So they are a lot more motivated, more challenged personally. And even before that, the fact that this is not a translation project. And it’s like linguistic consultancy, if you like it’s something that they don’t do everything. And perhaps a lot of them would like to do, they would like to have the opportunity to do that. One of the things that I hear very often is that, you know, guys, some of these questions that you have for my language, I have never had to think about, I don’t even know how to answer these questions, I have to do research for days sometimes to figure out how these things work. So that gives them a new kind of perspective and a new incentive to be more involved with what they do. And, implicitly also, they also realise that there are other opportunities out there, that there are other kinds of jobs that I could be doing with my skills, apart from just translation. And this is one of the things that technology has sort of allowed to grow in a more much more aggressive way, let’s say in the last few years, the fact that your linguistic skills are not just for translating, or interpreting they are for anything that has to do with language technology, is just that you haven’t had the chance to do that yet. But sometimes the chance emerges, like for a project like ours, and they get a completely different sense of involvement. So that’s why I think we get that kind of positive reaction. A lot of the times
Andrej Zito
yes, yes, yes, exactly. Yeah, you’re like the, I don’t know the hot chick that you meet in a bar once once in a blue moon, and then you go back to your boring. So so yeah. So that’s why I completely understand like, why the contractors, like working with you because you’re special, and you provide something different but but yeah. Do you? Do you by any chance? Give them some sort of a credit? Like let’s say I use lexica? Can I somewhere See, like who is behind the checks for let’s say, Thai language?
Vassilis Korkas
Or we can’t technically do that somewhere where we could say, we have only worked with one linguist for one language, because that’s very rarely the case. But what we do usually is we can provide, let’s say LinkedIn recommendations on their profile and say, No, this person has helped us develop the checks for that language. So you know, people know that they have done this kind of work, but we don’t necessarily repeat
Yannis Evangelou
recommendation letters to come together, get a new job or whatever. Yeah,
Vassilis Korkas
that’s has happened many times before. You know, this is not something that everyone necessarily even thinks about. But you know, at the at the end of a project like that, we always tell them that, you know, feel free to reach out, if you need that kind of thing. Some of them doesn’t need it, they don’t need it. Okay. But because it’s a kind of a different service as well, it’s not necessarily something that they need to prove to someone else in the future, at least very often, from a
Andrej Zito
different bucket going back to how you split the roles. So I understand that Mitra is the is the coding genius? Let’s see he was originally there for the languages. So explain to me, maybe this is a question for bacilli? Why is Jani is the CEO?
Vassilis Korkas
Well, yeah, this is, let’s say, apart from the, the progenitor of the idea behind like skull, this is not just an honorary title. He is the one who is basically right in the middle of everything. If I were to say that, Demetrius has a distinct role, and I have a distinct role, Yanis can do a bit of both. And plus, on top, there are things that you learn on the way about the role for us when we started, and we got together and started thinking, Okay, now, what title does each of us get? It was just the three of us in the beginning. The last thing we cared, we cared about was what title we’re gonna get. But we thought, okay, what’s the more natural, let’s say selection based on the skill sets that we each bring to the table? So Dimitri, is CTO? Fine. I have nothing to do with the coding. So I can’t be anywhere near any of the technical roles. So what’s the more generic one? Operations? What’s left? There you go. You have Yes. But
Yannis Evangelou
just operations. Yeah. Because bacillus bacillus. In fact, he acquired another skill, which had to do with finances, meaning that bacillus is actually also the CFO. So he does everything which has to do with the financials of lexica. I mean, we don’t have like kill slash CFO slash CMO.
Vassilis Korkas
So I’m not doing away
Yannis Evangelou
with do obviously typically,
Vassilis Korkas
but
Yannis Evangelou
but he did. So whenever. At the first level, whenever we have to do anything, which is about financials or legal stuff, it’s always about steel is who has the first take on it, Dimitri son died, we have no clue about it. And as you know, in our original, originally, we were only writing something else. Such as co founder, were three co founders. I think one of our advisors said, Look, guys, this is ridiculous. Like, get over it. Just get get an acronym. A currently mind yourself, and, you know, become a bit more mature. And yeah,
Andrej Zito
among the three of you, are there any weaknesses that you don’t cover? Like you as a group of three is or weakness that you have as a group of them.
Yannis Evangelou
And main one has to do with how reluctancy to expose ourselves like this is common across all three. Whenever it comes to interviews, presentations, conferences, and so on. I think all three of us are really okay. Really good probably at be having one to one conversations with people. Not talking business now. But getting to know someone and I think this is something all three of us are probably quite good at, but just put us on the stage and ask us to present something Oh, I suck for sure. Demeter is even worse, or even worse than the Phillies can do it very well, but he avoided meaning that he’s obviously much better than us. But he also avoids it like doesn’t want the light on, on the time.
Vassilis Korkas
I used to do it no, we don’t want in the past, but you know, I kind of got over it. You know, for academic conferences, you had to do it all the time. So I was there, I was out there a lot. But, you know, for, for a professional setting, it’s It feels very different to me. And I will always be more reluctant the same way, for example, that all of us have a hard time going out and about for sales. Like, our business development is basically only inbound. If somebody comes to us and asked us, you know, can you tell me about lexica, we can tell you everything. And it might work. It might not work. But outbound sales are basically non existent for us. It’s
Yannis Evangelou
called the males are called Khalsa. There’s nothing like that. There has never been anything like that. And yeah, even if we’re talking about really large constellation buyers, coming very close to us, and then disappearing for some time. And we were not pinging them at all. And they were getting back to us, let’s say two years later, and they were signing. But I think they were also surprised. Definitely not impressed. But surprised by the fact that they were coming so close. And then we were not even sending an email for a couple of years or whatever. Like, okay, what’s wrong with those guys? It just that. Within that, you know, if, if the company comes, if a company reaches out to you, and they know, they reached a stage where they know what they can get from you. They know their expectations, they know what you can offer. They will probably get back to you once they feel ready, or they will never get back to you. You don’t need to pick them every week with a week with cold email, cold calls or whatever. There’s no need to it does work this way too. I mean, this will work, but doesn’t work for everyone. It doesn’t work for us.
Andrej Zito
Right? Yeah, I feel like I’m going through the same thing like with the with the inbound marketing, like I’m not doing any sales, I don’t go to conferences or anything like that. So in order for that to work, your marketing needs to be somewhat there. How would you rate your marketing call a scale of one to 10?
Vassilis Korkas
From a scale from a scale from one to 10? We cannot really, because
Andrej Zito
it doesn’t fit the scale.
Yannis Evangelou
To reach one at some point.
Andrej Zito
We’re not yet there yet. We’re joking about it. So explain to me if your marketing is at level zero. why do companies? How do they still find you? How do they contact you? Is it just word of mouth? Is it references? How does it work?
Vassilis Korkas
A lot of a lot of it comes to the websites. So we did do some SEO work some time ago. And it seems that every now and then one of them will crop up. I think also
Andrej Zito
there you go, it can be zero.
Vassilis Korkas
Last year, it was one for a few months, and then it went back to zero. But I can’t put it all on that I think the needs of many companies are changing. And they are being more proactive in how they search for different services. And when they say Google, I don’t know quality assurance, or L QA for translation or something like that. We will be up there inevitably. And when they realise that our platform is entirely online on the cloud. If that fits their infrastructure, then they will be interested so they will reach out. And that is a lot more common these days that this kind of profile matches what we offer.
Yannis Evangelou
There. I think there are also a couple of things. On top of that, one has to do with a comparison study we did together with Mimsy. Years ago, when we were comparing our way, and they were comparing lexica with other QA tools, but also with the QA engines of established TMSs. And we still, typically even two years later, they get back to us and say, you know, what, we just were named partners, we just read that report. And, you know, we have exactly that issue, can you help us solve this? So this one thing. And the other thing is what you mentioned, Andre, which is the word of mouth, meaning that, especially in our industry, it’s also about your goals, do you want to sign 1000 Small to medium LSPs, then this will not this will not be the word of mouth or whatever there you need SEO, marketing, and so on AdWords, whatever, a lot of integrations and so on. In our case, we only target translation buyers, we have three, four LSPs. But our main goal has always been to sign translation buyers. And this our focus, and translation buyers, you know, they meet each other, like those managers meet each other at different level quite often. And they say, You know what? We’ve been using lexica for the past six months, it has worked fine for us, why don’t you try it. And this is how it works most of the times. So it’s not an excuse for the lack of our marketing right. Now, we were also thinking that if we overdid it with marketing with SEO, what did SEO bring that it brought us a bunch of small LSPs for which we would need to spend spent a lot of time in support to put the other way, somebody signs a contract of 500 euros a year or something, but you need to spend 100 hours with them at the end of any year, right? It’s better if they don’t find because you will never you will never do any other development work, or whatever. Because of that support. While in large translation books or large transition buyers, what we’ve noticed is that not they have their own development, development teams. And the what they expect from you, in terms of support is mostly reliability. If your API is reliable, if what they get back is reliable, and you are also responsive. That’s it. They might not even email you for for a year. And you might think, Oh, they’re probably not happy with the service. And then they are like, oh, let’s, let’s renew for next couple of years. We’re very happy with it, you know, we’ve never had any issues, for example.
Andrej Zito
Right? So I guess poor marketing is a way how to qualify your leads.
Yannis Evangelou
Exactly, exactly.
Andrej Zito
But explain to me the whole idea with with your, I don’t know, business model or concept, why translation buyers, and if translation buyers doesn’t mean that the translation buyer needs to have their own platform where they handle translations, so that they can integrate and why why isn’t it better for you or maybe even more profitable to integrate with TMS because with TMS basically the TMS does the selling for you like they have to find a row cleanse and you’re already part of the the bundler package, can you somehow explain this all one?
Yannis Evangelou
This is an excellent question. And I have to admit that I was like okay, he will probably ask you some points, you should probably ask it at some point. Because, we we follow this model that you described with regards to a TMS integrating lexica and then kind of either risk either using it for across all his clients or reselling it to some of his clients or whatever, you know, we have this we have variations of this model in our integrations with makeup, trans effects and crowding. And now it will also be easy look, TMS, but either link, so there will be like the fourth integration working more or less this way. But when it comes to integration with phrase, SDM, localised and so on. We try we injected with so we built a Chrome extension for all of this for free, in which we were injecting our live to a side panel, meaning that you type your translation crazy, extreme or localised. And there’s small side panel and bottom panel most of the time where you could see or highlight highlights like grammerly, in a way, it was like a clone of your target segment where you could see all results, and you were fixing everything according accordingly. However, none of the three were interested in taking this further. In this case, we would have to build a much more advanced Chrome extension. In fact, we would have to port pretty much the whole lexica to a Chrome extension. And lexicon now is not like like stick versions, 2016 or 2018. There are a lot of servers, there are a lot of NLP modules, which reside in their own server. And so there are a lot of things that you cannot do, even if you put them to Chrome extension, this becomes just too much, that was the fun thing. But even if you do it, the thing is that, then you also have to support to do extra UI support. Those people change their UI all the time. So do I, even once Lexus has changed, on their end, can break the integration. And integration does work in the back end, you don’t see it in a way and you have to work on it all the time. But it can be frustrating. So in order for us to advance an integration with within the TMS, our agreement should be like, with transfix, made cut, crowding, and so on, meaning that both of us do the integration together. We help them and they help us. It’s not like us, coming from the outside trying to inject code in with various hacks. Because a Chrome extension, what is it at the end of the day, it’s like a little hack. It’s an official hack in a way, right? This is not how things work. This can only be like an MVP. And this is why we build them back then it was like lightweight parts of lexica. So phrase exteme localise. Are you interested? Should we, we don’t really have a situation we made those integrations a bit deeper. But we need your help to do that. And your help doesn’t mean like six months of engineering work and so on. It’s just an API, you know, there’s no UI to integrate. But they didn’t, so we didn’t take it. We didn’t insist on that. You’re right to say that this would have brought a lot more clients, like, you know, having an integration phrase and being able to sign hundreds of clients at once. Just like that would have made things much easier. But what we notice aundre is a lot more companies can build their own TMS. And either by using open source, like installing, like, either by using private installations of open source tools, like private installations, make them and then asking for an API key or building kind of basic clones of some TMSs without being overcomplicated. You know, the cut tool doesn’t need to be over complicated. A cloud based tool you can use even memo queue, they have this web based cut to market trance I think, or whatever, which is a lightweight version of memo queue. But for translators, it’s fine. You know, it’s it’s enough. But a lot of people
Vassilis Korkas
invest maybe not when you’re laughing, maybe not.
Yannis Evangelou
Yes, All right. Yeah. Let’s let’s say that it’s not enough to say that it’s enough for waiters, let’s say it’s good, it’s good enough. So, a lot of people develop their own versions. Also, because they can control data they can. One thing has to do with privacy, another thing has to do with cost. Meaning that even if they have a lot of money, they can spill into Let’s pay 300,000 a year for a lot of licences on a commercial TMS, they prefer to spend those 300k In two years based engineers, for example, 150k each, who will be maintaining that TMS internally, and nothing will be going out, and it will be there, and they can do whatever they want. And I think what quite a few clients have told us is that they get frustrated with the analysis are, they don’t really have it, like, okay, it’s pretty much the same, you know, a weakness and strength, we cannot really decide, but it’s not what we want, I mean, we will pay that much money for this, we will not, we will not be that happy. At the end of the day, we will not have any control of anything. And they will. So they also change their conditions. It’s not just about pricing. Sometimes they say, for that price, you get five feet and unlimited work. And then you say that it’s not about unlimited volumes, they have assigned a certain volume of words to every seat, but they say it’s unlimited volumes. It’s not like, you know, a lot of things like rock, which frustrate translation buyers, and they say, Well, you know, we can take control of that. And this has become easier in the last two, three years. From what we’ve noticed.
Andrej Zito
I don’t know how to ask this the best way, but when I was reading about lexica, the first thing that definitely struck me was like it’s epi only. And I think I’m like from what you’re sharing, I kind of like get the sense, but also at the same time you’re comparing it multiple times to Grammarly, is there? Is there ever a plan or like a thought in your mind to I don’t know make lexicon like develop the UI and make it accessible to let’s say a regular translators, somehow,
Yannis Evangelou
hey, we did, like 50 demos in six months. They were not they had no intention at all, to add extra steps to their workflow. They wanted everything consolidated. And the only thing we’ve done is that we’ve kind of transferred the light to a to the API, meaning that if anyone needs any kind of help, in order to kind of highlight everything, the Grammarly way onto their editor, which helps them do that via the API. So we have transferred this a part of that on the API. But yeah, I think we have a an environment which could, like 10 years ago, this UI would probably be the next logical step, like moving away from excellence verificar K distiller and so on to a cloud based QA process, K workflow, but But yeah, people want everything consolidated. That’s nothing. And at the same day and trade, quite frankly, people often often have expectations which are not realistic. Meaning that I remember back in back when we started, a lot of the companies we were talking with, were saying were like, oh, so do you have any AI? So do you have any anything like auto correction and so on, and they didn’t have any issues which would be solved using auto correction. They just wanted to have something impressive in a way and we started as a rule base to a tool local specific role, basically a tool which became a hybrid then, so a lot of roles with a lot of machine learning at the same time. So a hybrid because this what we also find, to be working are better when it comes to quality control, like, not a blackbox of a neural network that you cannot really interpret, and not just rules because language is not just about not a set of rules. And you also, you cannot really maintain a system which contains only rules, the rules have to add new rules, or now have to use realise that, let’s have another if statement. So it’s an infinite bunch of if statements at the end of the day. So we came up with that hybrid. However, quite frankly, if we have kept the UI for our demos to this, where we show lexica to the run, we tell them that, you know what, we think that if you if you implement QA, or, or a separate click View, on your own platform, maybe this is approximately how it should work. Like using filters, which lead to a common history. So also everything could consolidate in everything you’d like to AU. And most of them have done at the end of the day. One of them didn’t. And two years later, they came back and said, You know what, we’ll do everything from scratch, we’ll apply the view that you have the shared with us. Just because we had spoke a lot about it in the first place. Or you remember when the beginning of our conversation, I said that, in most QA reports, there’s a difference between what we call thinking fast and thinking slow. So if you have to read the description of an error, you think slow, you have to process the things. While if you start using something like colour coding, it will take you a few minutes, hours, days, or whatever marks, you have 10 categories, let’s say with 10 Different colours or strikethrough. So underlines or whatever, then you know that, okay, so red underline is a terminology error or strikethrough. It prohibited word or whatever, blacklisted word. So you get used to thinking fast, but who wants to think slow, especially if you’re a translator, or reviewer. And this is a time that you spend without getting paid, don’t get paid extra to QA, your work, you see a bunch of false positives, it’s a long list. And you have to go through everything because there’s a block here, you cannot submit your project if you don’t go through everything. And also tick everything you know. So what do most people do more often than not, you just check old boxes or something. And they ignore what the report shows. But it’s also how they make things right. There’s always that thing that you missed.
Andrej Zito
Okay, two last questions. You mentioned that lexica supports now. One 160 plus or minus 150. And you started with four, how not mistaken for six? Based on what need decide which locale to add to lexica
Vassilis Korkas
150? lookouts. Yeah. Yes. This kind of changed over time, in the beginning, we had a portfolio of what we would consider essential ones. And we spent the first, I guess, two, three years without thinking about what is a priority, because we already knew what the priorities are. So we looked at distribution, what languages had more locales that were spread in more markets. So like Spanish, for example, we had to think about Latin American, Central American and so on. Portuguese, the same thing. We had to think about Brazilian Portuguese as well. But later on, it started becoming more an issue of aligning with what our clients needed more. So we started getting more requests about locales from Asia from Africa. The last big bunch that we did last year was 16 locales from Central Asia, Southeast Asia, and Africa. And all these splits are low resource low cost long term loans. So it always takes a lot more effort to build a locale like that to build QA support for it and spell checking. And we have to prioritise that well first Some languages, it’s also more difficult to find the right people who have the background you can trust, they will be able to help you. But for a lot of them is also the fact that they are just extremely complex languages. Grammatically, they have different norms and different standards. There are cases, for example, where, let’s say in some Indian languages, there is extensive transliteration. And there is no standards for transliteration in any of them. So anybody can do whatever they want. Or there are other languages that have noun cases like Greek and Czech, but you have, I don’t know, Georgian that has 10 different cases, and another 15 post positions that change how a certain word, we’ll look at the end of its transformation. So getting all that coded into a spell checker, for example, is madness. It’s extremely challenging. And it’s also something that requires a lot of time to collect the data you need initially. So it might take weeks before we have everything that we need, for all the checks that we have. Because the idea is that whatever checks we have, they will work the same way for every lookup, no exceptions. The exceptions are that there might be locales that have a requirement for checks that don’t exist in other locales, there is, for example, one case in Thai that doesn’t use any punctuation at all. So you will have an English short segment with a full stop at the end, Thai wouldn’t have a full stop, because it doesn’t need IDs. So, and no spaces. Yeah, I mean, so any check that we have for spacing and punctuation, for tie has to be custom made. And the same goes for every language now has its own peculiarities, we have to take that into account.
Yannis Evangelou
And you find that you’re dealing with in Canada, right? And you decide to assign them to job. And then you realise your own these more like the third language? Are they not exactly native speakers of girone? So the parents were keen on the native speakers, but they’re not they’re like, let’s say like my kids who might be speaking perfectly, but don’t write perfectly or well or at all post something. And then you say, Alright, I will cross check with Kieran, the speaker from and then to find the current speaker who, whose, whose native whose mother tongue is not clear on the they say that it is guilty, but it’s not clear on it might be Hausa and Kieran days the second language or Swahili and given this second language and then you have to cross check with another one. And this is something like a vicious circle. We have had issues like that in the beginning. Not for God not for the for other languages. No, no, no. No, not that he was excellent. But we learned the hard way meaning especially when when we started dealing with Indian locales in which there is no standardisation first, and people might be talking at least speaking at least two Indian locales, like being flown the fluent in Hindi and Marathi, which are very similar languages, but similar doesn’t mean the same. So asking from a Marathi language to provide you with data for Hindi or the other way around, doesn’t exactly work. So we’ve always had like this. After seven stage, we we’ve always had the security layer of at least two linguist per locale. So that we cross check what we kind of validate what they give us back and it was turning out more often than not that some people were just giving something different back it was not that language was a different language or a mix of languages. So the lack of standardisation is one thing, finding reliable people either in country linguist or linguist who have moved somewhere else in Europe, let’s say or Canada and haven’t been in touch with the actual language for a long time. So they say, Yeah, we do speak that language, but they use the speak that language not exactly the same, right? They’re not in touch with that language anymore. To comfortable with the human resources, though, so it’s low resource, not also, not just in terms of data, but also in terms of human resources. So challenging, and then has to do with the actual structure of the language. For many of these. There’s some some no standard, like, yours might be like, Oh, we use apostrophe all the time for that we separate this all the time. And others say, no, there should be no apostrophe, like there are multiple schools of grammar. And this can be very challenging. But what we’ve learned the hard way is that in those languages, you cannot really go for perfection. At the beginning, you know, we start with a bit lower expectations, and you grow. But you have to update your data, you have to update your models, your data, your rules, everything all the time, based on the feedback. And I think we’re at this point where we feel very confident about our checks in goes, flow research, look, locales do you want, meaning that, you know, it’s very easy to find reliable linguists and a lot of reliable data in European languages? Try to go out of your way to get out of Europe and find reliable data lab linguists, and so on. There are but a linguist, yes, you can find but you need to pay a lot for the library ones. And you need to find them because everybody wants linguist. So when we were coming up with a language that we wanted to go with us, it was turned out that he had just been hired by Amazon, Google and so on, because that was a good language to language.
Andrej Zito
Yeah, okay. Final question. For Yanni. What do you think is wrong with our industry?
Yannis Evangelou
I didn’t expect that.
Andrej Zito
Way, what if you actually know what to say that maybe but certainly can do it? I thought it actually no. You love him roast, okay.
Yannis Evangelou
With a baseball face? Why he says I’m a tornado. That’s why
Andrej Zito
specifically, I asked you
Vassilis Korkas
to come in and say do you want that for basically?
Andrej Zito
Can you maybe pick one, which you think is like the worst one?
Vassilis Korkas
No, no, no, no, no, no. I was just joking.
Yannis Evangelou
First of all, I think when it comes to technology, our industry compared to almost any other industry is way back. And this means that every time there’s a trend, hi, whatever, we feel that we’re way ahead. No, we’re not, we’re still way back. Right? I feel that a lot of our companies, and that might even include us would not exist in 2023 in our industry, like lexica phrase or put whatever you want in that bag of brands. We shouldn’t have been there 10 to 15 years ago, 10 years ago, at this point, we should take a break, we should come in there five to 10 years ago. When, like five years ago, when everybody started talking about NMT. And now everybody started talking about chargeability even for chargeability you see that people are now doing the six fields that this has to do with our industry like the localization industry. Now we can be affected. It also involves us and so on. But it is not an achievement of the localization industry. It comes from the outside. So that’s one thing. The other thing has to do with the fact that I feel that a lot of a lot of LSPs have actually done so the traditional industry of options does more harm than good? And why? One thing has to do with the fact that they transformed the translators profession into something which resembles more to factory job. So why would someone wants to be a translator nowadays? And if you want to be turned or become a translator nowadays, for how long? Do you think you can be working for an LSP? Let’s say, not that much. So the burnout is, is there, you will have it one way or another sooner or later. And the other thing is that the last thing for me, at least, I think there’s a list of things, but then try to finalise my words with this one. Yeah, I will. I will, I will, I will. Probably the fastest way to keep it short.
Vassilis Korkas
Yeah.
Andrej Zito
Are you thinking about like a politically correct way to say it or No, no, no, because I don’t
Yannis Evangelou
have that issue, at least I’m not correct. At all, at all. So no worries about that. So the thing is that our industry is built look, a lot of networking, it’s less, it’s less product based, it’s more network based. Meaning that even if you don’t have any product, let’s say that you have an idea with no product, do you have a process with no product at all, or whatever, but you have a strong network, you might achieve much more than someone with a strong product. And this can be frustrating for people who start building products, we invest a lot of time and effort for problem solving. So what I notice, and this is not a personal frustration, it was a personal frustration years ago, it’s not anymore. Just to be clear. A lot of us tried to solve are in this industry, just because we want to solve or we try to solve some problems in what we love, which is languages. And we noticed that a lot of people in this industry have nothing to do with languages. They’re here, because you know, they’re their managers, or whatever they used to manage other sorts of companies, but they realise that they can also do that in languages. They don’t care about languages or problem solving in languages at all. And this affects the whole cycle of our industry at the end of the day. Because we often have a bunch of managers who know nothing about what is going on. And they try to decide job, nothing about the product technology, the actual work in our industry, and they become the decision makers, or even the references. And they tend to ignore those people who put all quit all their effort, time and money in building products, which will solve the problems, some of the problems in the industry. So at least those they try. Don’t really don’t ignore them. This is why I have at least one principal, whoever writes to me on LinkedIn, or sends me an email or whatever. And they say, You know what, we are building something. Do you want to see it or whatever? I always say yes. And I will not do any name dropping. But seven years ago, when we were starting, I wrote a lot. I my advisers told me, You know what, you will have to write a lot of cold emails, like find those people that you want to ask can write a lot of cold emails. So I wrote a lot of cold messages on LinkedIn and 99 of them like no responsible, no connection, no responses. But you know what, there was a guy who is the CEO, the founder and CEO of a very important language technology company, and invited me to his office next day. Come here we’ll discuss I will share my experience and so on. And we’ve met a lot of times since then, has helped me a lot. I haven’t given him anything back. He doesn’t have anything to gain from lexica and so on. But for me, This guy has served as an example. Like, okay, this is how I wanted to be treated. And this is how I want to treat anyone who would be in my position and in my place and just send the cold message in like, oh, you know, I’m building that. Can you help? Can you just? Can we just talk about it? Sure. Anytime,
Andrej Zito
right? What’s his name? It’s a good. It’s a good example. You don’t want to name someone? No.
Yannis Evangelou
Because I haven’t asked for his permission. But yeah, he’s a good. He’s German, German. Very good.
Andrej Zito
Going back to where we started the whole interview. Right. All right. Um, my final question for Vasili would be how do I ask that? If you could speak to the minds of everyone in the industry, what would you tell them?
Vassilis Korkas
I think if I wanted to keep it as a short and catchy message, it would be don’t be afraid. I’ve realised over the years that both when we are at first few steps of our career, whether as learners or professionals but also later on, you can see it with more experienced people in the industry, that they’re afraid of change. So I would say don’t be afraid.
