[00:00:02] Speaker A: I've been attending the American Translators association for the last 15 or so years. One of our colleagues within the industry that attends that regularly is Jay Marciano, who's been a huge evangelist for machine translation. And I saw him, you know, twelve years ago where people were saying that he was evil man and he was just in horrible ideas and it was completely dangerous what he was just. And then I saw him last year and people were coming up to him and their attitude was, thank you. It makes sense what you're saying. We need to get grips with this. When they were talking about machine translation, they were generally talking about Deep l and they weren't looking at anything else. The translators, they were very open to what AI could, could deliver and how it could be used. And it was not at all the image that had existed ten years ago.
[00:01:01] Speaker B: The following is our conversation with Peter Reynolds, the co CEO at Memoq. We talk about how he became that co CEO that he is today, his relationship with clients and customer success, and how, when and why they decided to enter the german market at a key moment in Memoq's history.
Thanks for listening. This is localization today.
[00:01:30] Speaker A: I've been involved in the translation industry for about the last 25, probably 30 years at this stage almost. I've been involved in it through quite a lot of interesting projects. I started my work in a translation company that was called Soft France. And they got the name Soft France because they had this idea of selling translation services to software companies. And they were probably the first pure localization company out there.
They got sold to Berlitz, and Berlitz got sold to bound global solutions, and bound global solutions got sold to Linebridge. And a lot of the ideas from this small company in Dublin, Ireland, which later became quite a big company, are still being used by Linebridge today. And that, I think, is quite interesting. The company had an online portal in the last millennium, around 1997, and it was one of the first companies in the world to use web services for a commercial transaction.
We also were instrumental with partners, particularly in the large software companies, in creating a working group that became XLIF. And it's probably the dominant standard within our industry. And it started off with a group of people that just got the right people together in the room and told them, come up, do something clever. And what they decided to be. Big problem they had was interoperability and file interchange, and they set out creating a standard that deals with that. I think they did.
I've been involved in technology pretty much since the beginning of my time with this and around 2005, I left Linebridge to join a company called idiom. Lionbridge was based in Walter, Massachusetts. And while I was living in Dublin at the time, my headquarters of the company was about 2 miles away and also in Walton, Massachusetts.
So I worked for idiom for a couple of years until they got acquired by SDL. And at that stage I became part of Mama Q. And the company was starting off and had all this growing pains of a, a young company. They were based in Hungary. Nobody had heard much about them. And we tried to figure out how to make ourselves known, and we eventually we did that quite quickly, and we spotted that translation companies at the time were not being catered for, and we decided to cater for them.
And it was quite led to quite a lot of success. And we've since grown a lot, and as we now cater for a very broad sector of enterprises, translation companies and translators.
[00:04:44] Speaker B: Great, thank you so much for that background. And before we get into the approach of how you're catered to companies in the broad sector and translators, let me step back a little bit and tell me about your relationship with technology and what was it like to introduce technology to localization in 1997 versus what it is today in 2024?
[00:05:13] Speaker A: Like you had a lot of, you had quite different challenges at that time, 1995, when you localized software, you would probably have to deal with real, you know, you dealing with the nitty gritty. And I remember once I saw what I thought was a mistake in the Microsoft product, where they had double a at the end of an order of words, and it was in Norwegian, and that's correct. But I reported this as a mistake and they fixed it probably, you know, it's now 20 years later, so I can talk about it, but it probably caused them some problems.
[00:06:03] Speaker B: Very likely.
[00:06:04] Speaker A: There was the same sort of attitude, though, then as now, in that everybody had wide eyed curious, they wanted to figure out how to improve things. The core idea what, what translation companies wanted to do then and what they want to do now is to try to improve the efficiency, try to have it so that you will improve the quality, but do it faster and at less cost. And that's what enterprises have been trying to do. That's what.
And it's all just been the same sort of motivation, constantly trying to improve the quality and improve the efficiency and the value that's offered, you know, by our sector. And I think that tells a good story about the people that make up this industry.
[00:07:00] Speaker B: And that's absolutely correct. Of course, we're talking about the people and the first group that I want to talk about, when we talk about the people involved in the industry is your clients. So you talked about how you were catering to specific group of language companies and you also mentioned translators. How do you experience this focus? And of course now you are co CEO, so it's a little different from when you were perhaps executive director and you were doing other things in the company. You have a focus on technology, but how do you experience that focus on customers?
[00:07:40] Speaker A: Yeah, our customers are wonderful.
They're very, at this stage it's about 40% enterprises, 45% lsps, and ten or so percent are directly translators, although translators also work with us through enterprises and true LSP's and then there's others like government organizations and so on. But we've always been quite proud of our customers, that they've got strong opinions and you know, we've always liked the fact that they've been people that know what they're doing and they want us to create our software in the best possible way and also won't take, and you know, they won't take errors or issues with the software.
They're not happy and things like that. And it's quite a nice way for customers to be that. They want you to be the best and we'd like to be the best.
So our approach to our customers has always been one of absolute respect and we get an awful lot of ideas for them.
One of the things that we have at Memocue is an ideas portal where the customers come to us and they suggest ideas for the future of the software.
And it's often the case that somebody has got a brilliant idea and other people have, you know, say things that like, that's already possible or whatever. And it may be that, you know, they don't need what they think they need because they, you know, it's already available there. One example of this was a customer that suggested, could they have something where they can specify certain terms, do not appear in a find and replace?
And other people wrote back and said that, oh, you can do that already, just use Regex. And the first person says, wait a second, you can do that already. But I can't, I don't know how to do this. So we, our product managers sat down and talked to all these people and talked to a lot of other people about what they were doing with this. And we didn't build what he had asked for. Instead we built something a quicker way, a sort of wizard for rating Regex quite simply so it effectively does what he wants, but it's also really more powerful than that. And I think that's one of the ways that we've approached customers that's been very effective.
[00:10:30] Speaker B: Is there, and thank you for sharing that story. Is there any idea that have come from that ideas hub that you can recall that has been something fundamentally transformative for Memoq? Or is there also been a type of idea that has just been shared over and over and over and over that you've come to the decision at some point? Okay, let's get this done.
[00:10:57] Speaker A: You know, it's rare you get a magic idea that just changes everything.
One of the things that we're doing at Memocue recently is around the in country review.
If you're an enterprise, say, if you're a manufacturing company in the US and you sell to Germany and you sell to France, the probability is that you've got somebody in the german office of your company or the french office who has to sign off on your translation. Say, if you put an ad in a trade magazine, the person in the german office would probably sign off on this. And they almost always have very complicated processes for doing this. Like, sometimes they actually print off the sheet of paper with a copy and write on it and then scan it and send it back. Other times it's adobe with post it notes and whatever, but it's invariably very complex and that. And we spent a lot of time hearing about these ideas through this ideas portal and through other places. And we talked to an awful lot of customers. And what we developed is an in country review process that is remarkably efficient. And for example, if you were, say, your manufacturing company that I mentioned, for them, it saves them a huge amount of time and hassle and whatever, what would happen is that because this was a cumbersome and difficult process, when they approached a person in the german office, they would not like it, and they'd be figuring out, oh, well, I have to do something that they absolutely hated because they won't figure out any excuse not to do this. With the process that we've introduced, it's just much more efficient.
The in country review happens for, you know, seamlessly, and it's all built into the tool. And that's an example of something that for those customers, it's transformative because their quality process is essentially built around in country review.
[00:13:25] Speaker B: And thank you for sharing that. It's, of course, evident that once you're prepared and you have the right mindset, you are open to receiving these ideas, to making the changes that the company requires. And I, of course, understand that Peter Reynolds did not. Was not born like he is right now. He was not born co CEO of Memoq.
There is. There is a huge process by which you get here.
Could you tell us a little bit about your history, about your stories, and perhaps anything that prepared you to say yes to Memoq, to becoming, of course, initially executive director and then co CEO of the organization.
[00:14:13] Speaker A: When I started with Memocue, as I mentioned, it was a very small company. There were three or four people working for it.
All the sales would be done by someone who I think is a colleague of yours, ish Van Lengel, do everything himself. And he'd also deal with a huge amount of relationships.
What they said to me is that we've no money. Can we get you to do work for us, and we'll give you shares in the company? So I said yes, not knowing whether that would be very, you know, kind of maybe just some fun that I had with these nice hungarian people or whether it would be a successful business. But we turned it into a successful business.
And, you know, I've got some background. I worked in the industry.
I've seen what issues there. I've got an MBA. I've gone through business education. And I was able to put all that into practice within Memocube.
And my first role within the company was essentially designing and creating the strategy.
At the time I became CEO, we decided that we wanted a change of direction. We wanted the company to move faster. It was currently moving. And what we decided was there four owners of the company, myself and Balaj Kish, who's also co CEO, who's the other co CEO, were both owners. So we initially decided we would do this for a short time as temporary, you know, as temporary co CEOs. And people were demanding, you know, but inside and outside the company that we stay. And it was very enjoyable. I really liked the job. I liked the challenges. The people that I work with are wonderful.
You know, they make my job quite easy. You know, I just smile and take credit for everybody else's work.
[00:16:37] Speaker B: Did you ever think, was that, was that something that you had thought about and you had been CEO before, so that was something that you knew you wanted to do?
[00:16:44] Speaker A: No, no, I didn't realize I'd wanted to do that, but I pretty much enjoy it.
[00:16:51] Speaker B: Oh, great. And you mentioned you didn't know initially if it was going to be successful or not. You saw this hungarian company going through some trouble.
When did things started to change, and how did you realize things were actually going to work out?
[00:17:09] Speaker A: Around 2008, 2009, what we decided was that at the time, Tradus and StL were the biggest by far. Theyre still the biggest, but then they were almost a monopoly.
And Tradus is based in Germany, and you had a cross that are also based in Germany, and you had a company at the time, Entram, based in Germany. So the best of the translation technology was all coming from Germany. So what we did is set out to be successful in Germany.
And we found we came across a couple of wonderful salespeople who were german and were very happy to push this thing to try to be successful in Germany. And we found within a year, we had got, you know, Deutsche Post, we had got some really big accounts within Germany, and, you know, to win in Tradis's home backyard was indication we were going to be successful.
We also, at the time, Travis was going through a major change. They had a tool called Desktop Workbench, which what it was was add ons to Microsoft Word, and they wanted to go to a complete standalone application, and they developed a tool called Trada Studio.
And Trada studio was not compatible with Trada's workbench, but Memocue was compatible with Tradis studio and with Tradis Workbench. So we told people that we were more compatible with traddles than traddles, and that also caught people's imagination.
[00:19:10] Speaker B: Was this something you planned, or was it something that just happened?
[00:19:13] Speaker A: No, we planned it.
We planned it slightly as. And planned it, like, sitting around joking about and saying would be fun and then trying very hard to make it happen.
Like, it wasn't planned, as in, you know, it was planned, as in, come on, let's go for this.
And it worked.
[00:19:39] Speaker B: And you mentioned then this. Moving to Germany was critical at that moment because you were winning. Like you said, in the hunter of the leaders in the industry, was it always this rosy? Were there any moments of difficulties or challenges along the way?
[00:19:55] Speaker A: Yes, there's always moments of difficulties and challenges on the way.
You know, we were successful, and we had a clear idea of what we wanted to do. And then almost immediately, the other companies came along and copied us, and that obviously made things difficult. But it was all, you know, in retrospect, companies like Memsource and whatever did us a lot of good, because having good competitors helps everybody.
The whole competitive landscape has changed remarkably, and that's ultimately in the customer's favor.
It's much better for the customer, you know, be boring. If you were sort of so far ahead of your competitors that, you know, was just, you. You had nothing to do.
So I like the fact that our competitors are talented, very capable people, and I like competing with them.
[00:21:10] Speaker B: And that's really good to hear. I agree. Competition is very important for all industries. And you can see it in fashion. Right.
There is a huge level of innovation because everyone is allowed to copy each other, so they have to put their logos on top of their design. So that's the one thing you cannot copy.
But so innovation, of course, has had to play a role in memo queues.
Success and continuous success, I should say.
And of course, you have a number of solutions. You have technology embedded into what you do. Could you tell us a little bit more about innovation and the role that technology is playing today in memoq?
[00:21:57] Speaker A: Yeah, one of the things that Memocue did, which has, at the time we started doing it, we weren't sure exactly how positive it would be, is that we kept on reinvesting in our translation resources, like with translation memories, for example. We're the only tool that's kind of looked again at that and gone back to the basics and completely redesigned it for what we've released last year, TM.
And this is a good example of something where there was the benefit of actually doing it because we created a much more scalable, much faster translation memory system. But then with artificial, with generative AI, we found that having these resources made it much easier for us to come up with a solution around that. What we did when AI came out and we were listening to the same thing everybody else's, the first thing people were saying is that how this will change is that people will have a company will put in a prompt that says, I want a description of my car for whatever market or whatever, and have it sort of. The only difference is that they want it in five languages. So there'd be one prompt to five languages. And we thought about this and thought, well, wait a second, you could have done that at any stage because there was always a possibility of transcration, where effectively you have the idea and you send it to the people working in language and they create different versions of the same idea. That's what one prompt to many multiple languages could do.
And we talk, you know, the person that owns the content in the original language never likes this because they're always very nervous about what happens to that content if it's not a translation, that if it's transcrated, that they're always worried about how it will change and they can't control it and so on. So we don't exact same thing will happen to this, it won't be as popular as people were thinking of as being popular. The other realization that we had is that people were saying that all these resources that we have, like translation, memories and corpuses terminology, they will all be old hat because of generative AI. And it dawned on us that it's in fact the opposite. It's those resources that make things, that certain things possible. And what we did when we created Mamaq AGt, which stands for adaptive generative translation, is we get a segment, and then we firstly check the Mamochue resources, and we say, here's a segment to be translated, and here is the source and target for two source and target of some examples from the TM.
And what happens with.
You have the equivalent of effectively training a model on the fly. Because what happens is that the answer that comes out of it is accurate.
It's more accurate, and it's also using your style guide and using your terminology, because it's getting that from the in context information.
And we thought we were on something really big with this, and I think we still, we still believe this. But we released this around October, November of last year. And then around December, we saw this was actually what other companies were doing in different spheres, like SAP and IBM were talking about this, and they were calling it retrieval augmented generation.
And it was exactly the same thing. It was find some human knowledge. Adash has contextual information to the LLM, and you're getting back much better results.
And what we found is that if we have a, you know, if we don't use the in contextual information from Mama Q and we just use an LLM, the results are, you know, bad.
Usually there's lots of hallucinations, there's lots of, you know, things possibly wrong with it. But when we use the resources from within Mama Q, we're getting results comparable with any of the top NMT engines, and we decided to benchmark ourselves against the ones we perceived as being the best.
And I think we're onto something that's quite exciting here. And I think it's probably the future of how machine translation will go.
[00:27:38] Speaker B: And of course, you've mentioned a little bit, and it's very interesting. This month we're talking about the history of machine translation in the most recent issue that's going to be published in the next few days. Absolute.
Very interesting what has happened around machine translation. And some people say machine translation is the first form of artificial intelligence. And at the same time, of course, you did mention as well that some of your clients are translators themselves. And, you know, it's translation industry.
What is your current experience in the whole conversation? AI versus humans. And I know some people want to polarize it, but where do you stand in that conversation?
[00:28:23] Speaker A: Yeah.
Yeah. I think this is quite interesting. I've been attending the American Translators association for the last 15 or so years. One of our colleagues within the industry that attends that regularly is Jay Marciano, who's been a huge evangelist for machine translation. And I saw him, you know, twelve years ago where people were saying that he was evil man, and he was just in horrible ideas, and it was completely dangerous what he was just. And then I saw him last year, and people were coming up to him and their attitude was, thank you. It makes sense what you're saying. We need to get to grips with this. When they were talking about machine translation, they were generally talking about deep l and weren't looking at anything else to translators.
They were very open to what AI could deliver and how it could be used, and it was not at all the image that had existed ten years ago. I think translators were more nervous about statistical machine translation than they were are about AI. And I think the reason for this is that statistical information and machine translation was not as good as people were claiming it was. And, you know, people were making claims that it's going to totally change the industry and whatever.
I think your translators know that if you get to, say 95% perfect with AI based machine translation, that is really good. But there's one mistake in every 20 words, and that's not acceptable. If you're going to translate documentation for a machine that's going to be used in a medical operation, or it's not acceptable for an ad for perfume in a magazine.
You mentioned fashion earlier. Having one mistake in every 20 words for a description of some just wouldn't be acceptable.
And what people have been advising translators for many years, what they've been doing is getting more specialized. And I think that's the key, that when you're very specialized, when you're very knowledgeable, you've got a job. I think there are people within our industry making claims that are a bit wild, and I think people outside of our industry, many of them believe that translation as an industry will be gone a couple of years to be replaced by machine translation. But there was an interesting post recently from Scott Bass, who runs translation company in New York state, and he got a three page document, which was a privacy agreement that the European Union had published, and he put it into chat chibt, and chat Chibti translated the first four paragraphs and then stopped.
And in he said, note of to, you know, why, why won't you translate this? And chat GPT came back with lots of reasons as to why it was not the best option for translating it, saying that this is a very complex document, it's a specialized domain, it's a long document. And what, and what chat GBT suggested that he should do is firstly use a tool, you know, use tools that is specialized for that sort of thing, for translation. Because according to chat GPT, the segmentation will have less errors. And I think what they mean is that with the LLM type translation, if there's an error in second paragraph, it will be a much bigger error by the end of the translation. It will just continue with that. While if it's segmented and used in the way the cat tool does, that's not going to happen. The other thing that chat GPT suggests, that is use a professional translation service provider who employs professional translators, because they're the sort of people that can translate this sort of specialized knowledge. And I presume OpenAI are probably the leading expert in generative AI. If they take complex specialized documents, should be done by professional translators, then I think it rests the case. But there are people outside of our industry who think that translation will be gone very soon. I think we have to make the case that it's not quite as simple as that, and these things are complex, that we're all for progress in technology and improving the efficiency and getting higher quality at less cost. But you'll still need somebody to make sure that you've got that quality. And if you are, if quality is important, you need a professional translator, both in the process.
[00:33:59] Speaker B: And I think we're on the same page here, ignorance is bliss. And when you are outside of the industry, and I'm also an outsider now, being in the industry for a year and a half, then when you start seeing the complexity, then you start understanding why it's not likely to replace translators or specialized translators. Now, as you probably can agree with humans, we are great at using tools, and technology is just a huge proof of that. People start using them, playing with them, and memo queue is of course a huge example of this. How do you think specialization and specialized translators are going to evolve with the use of tools? There are specialized translators today not using tools, maybe because they are useless to them, it cannot help them. Like you said, a specialized translator using chat GPT not really going to be of any help to you as a specialized translator. What have you seen how is it evolving and where do you think it's going to go?
[00:35:01] Speaker A: I think, you know, translators will get more and more educated. They'll.
I mentioned the American Translators association at their conference was full of presentations about how to use tools better. I think every single CAD tool integrates with almost all machine translation providers. Translators do use them. They also talk to each other and know what they think is best. They know that if they're dealing with a certain language pair, a certain tool might be better.
We got feedback from a translator that they had started using Mamiq AGT and had a 400% increase in productivity. I'm not saying that they'll always be that much, but it was interesting that that's what, what they were saying. And it's a case of, you know, they used the tools and the tools allowed them to do more for less. And there's all sorts of ways that translators can use tools to improves their process. Not just with AI, but many of them have used voice recognition.
One of our customers for many years used to have a keyboard. Well, he didn't have a keyboard. He had what looked like a games console and headphones. And he did all his translation through true voice recognition.
And he was one of the top translators for patents from English to German.
And he would be translating 8000 words a day, while if he had typed most, you can type but two and a half thousand words a day. So he was doing a lot more words. And because he was a top translator, his words were not cheap.
You know, they hired by the best companies and whatever, and, you know, it's an example of how they use technology to, to improve their business. And I think it's something that we all should do. It's something, you know, we at Mama Q use various tools to how we improve creating software and running our business. Translation companies use tools like Mama Q to them to manage their process, to manage their quality and so on enterprises users quite effectively. One of the things that enterprises are doing with Memocue is they will have a copy of Memocus TMS and have a Memoq tms to Memoq TMS connection with the agencies working with them. And I think the old have to see this in a positive way. I think the tools should be enough so that people using them feel good about it. You know, they feel they get some fun out of using a tool and they can see the improvement and they. It's not a case that they're using a tool and it's like climbing Everest every day.
You know, using the tools should be like a walk on the beach.
It should. You know, we as tool makers have to make it so that they can see what they get out them and make them easier to use. We have a lot of time and energy into educating them.
And one of the things that we've been doing for many years to do exactly that, the Mamma Q fest conference that we're going to have this year in June, at last year's conference, the big question was, this was June 23, when everybody knew AI was going to be a big thing, but nobody knew what we were exactly asking that question. We had people like, for example, we had Jay Marciano at the conference. We had other people talking about. We even had a competitor.
Yves Champollion was talking about his great grand uncle, Robert Champollion, whom deciphered Rosetta stoned and was, you know, a lot of our knowledge about ancient Egypt comes from his work in translating, you know, the Rosetta stone. And he used tools to help him do that, that were techniques, processes, such as concordance, checking and whatever.
Some of the things that Robert Schampollion used to Arjan Francois sampolion, I apologize, used to translate the Rosetta stone are still used today.
I think we just have to be positive about technology and we have to be in control of it. It's not a case that technology is going to control us. We have to control others. It's there for a tool for us to learn from.
And the more positive we are about it, the better.
[00:40:24] Speaker B: Thank you, and I'll get into the memoqueue fest in just a little bit. But I wanted to ask you, memoq adaptive generative translation. As an AI based translation automation technology, where do you see it standing right now in the landscape of tools available? And where is it going?
[00:40:48] Speaker A: I think it's the most exciting tool out there that anybody's got. I think if any other of our competitors would like to say that something they have is more exciting than MMQ AGT, let's argue. Sure, we can find somewhere to have a discussion about this or whatever, but I think it's one of the most exciting tools out there. And the reason for that is that it's using AI. It's using llms in a way where we're effectively training a domain on the fly. If, for example, you were a big life science company and you wanted to use NMT, your first process would be to get all your translation memories together and to train them so that they be added to some NMT engine. And it was a difficult and expensive process.
What is happening with Memocue AGT is that happens on the fly. You have the translation resources in Memocue. If you don't have the translation resources in Memocue, we've got a live docs corpus that allows you to automatically aligned source and target segments and that can be used to effectively train the engine on the fly. So you've got the domain right without doing any of the hard work. And that makes it very, very exciting. And we're also at the beginning with this.
We're currently using GPT 3.5 turbo with the Microsoft Azure platform, but these companies are improving how yellowms works.
There's a lot of work to be done to make things, to have less hallucinations, to have improvements. As the llms improve, the quality that we get out of them will improve. So we've got something that's now very exciting, but it's going to greatly improve as time goes on. And I think, you know, that's an interesting something.
Good.
[00:43:19] Speaker B: And Peter, thank you so much for your time. I just realized we've been talking for a long time, and time just flew by. I really didn't realize. I have, I have a daily with the team and I just didn't, didn't realize. I was like, wow, I can't believe time has flew by so quickly. Before we go, could you tell me a little bit about Memoq Fest? What excites you about this year's version?
[00:43:42] Speaker A: This is something we've been doing for a long time, and we originally did it too, because we were a tiny company. Nobody ever heard of it. And we used Memo Q fest to give the impression we were much bigger than we were. We actually got called Mamocue at the first Mama Q Fest. We were called Mamoc until that. And then all our customers said, no, we're called new Memoq. So that's where the name, where the pronunciation comes from. This year's Mami Qfest should be quite brilliant.
We have a friend of Mamo Q who's been at, I think, three Mama Q fests, and he's always been brilliant. A guy called Mike Dillinger who has been someone who's leading how we deal with machine translation. And he's been one of the first to have to use machine translation for large companies in a way that was very effective. He's now putting a lot of his time and energy into knowledge graphs, and what we're hoping is that he'll be talking about how knowledge can be used within this setup. We've got other speakers, Gabor Ugrey, who's the person at Mama Q who came up with het, will also be talking at the conference.
The main thing, though, with this conference, I think any other conferences, is that it's not so much the person in front of you on the platform as the person beside you. You know, who you'll meet at the conference dinner or at the social occasions, because usually some, and it's always the case at Mama Q Fest. You'll find some of the most knowledgeable people within the industry and also, you know, the most charming and funniest people there. And it's that sense of friendliness and knowledge and curiosity that you'll find at Mama Cube Fest that makes it really worthwhile.
[00:45:46] Speaker B: All right, thank you. Thank you so much for that, Peter. And before we go, do you have any final thoughts, anything you want to share?
[00:45:53] Speaker A: Just to say thank you. I really enjoyed talking with you, and I've noticed the time, too. I thought we were going to talk for 20 minutes, and it's 50 minutes. So thank you very much, Eddie. It's great to see you. And thank you.
[00:46:14] Speaker B: And that was our conversation with Peter Reynolds. Remember that this was multilingual. You can find more on our
[email protected]. You can also subscribe to our magazine to receive it at your doorstep every month and on your inbox every month as well. My name is Edi Arieta, CEO of multilingual magazine, and this was localization today. Thanks for listening.
