From Machine to Meaning: Malt’s AI-Driven Localization

Hello, welcome to Localization Today. The show where we talk to the people driving innovation in the language and localization industry, we are looking to explore how language professionals make products and services resonate across cultures and markets. My name is Eddie Arriotan, the CEO here at Multilingual Media and today we are joined by Teresa Torón Localization Manager at Mold with over a decade of experience at companies like Workday, Clue and Now Mold. Teresa has helped build and optimize localization workflows that connect global audiences. Well, strengthening collaboration across product, marketing, legal and customer teams. She's a champion of lean teams, smart processes and human focus localization Teresa. Welcome and thank you for being here. Thank you, Eddie. Well, I couldn't have that better thanks for that introduction. I think when you pair human intuition, knowledge and artificial intelligence, you can do amazing things. I'm a big believer that artificial intelligence is here to make us do amazing, beautiful work. So I hope today we can get some of your perspective on how you do it. Teresa, of course, how is Mold, how is your role, how are things going, happening and changing so far? So I think, oh my god, everyone is caught by this AI hype. I'm going to call it hype, but that doesn't mean I don't believe in it or anything. It's just really, it's like this big train that is just like going through when everyone is trying to get on it. And it's exciting at Mold. But you know, I started two years ago and so a lot of things have changed since I joined two years ago. When I joined, I was of course looking to automate and make processes more efficient. We work directly with freelancers for our translation so that that hasn't changed. But we have changed a little bit the way in which we work with content, how we automate processes and the role of the translators itself is changing. So yeah, I think, but I think I see that as something exciting. It's not, I haven't perceived anything bad about it so far. And then of course you have the AI trend in the product itself. So we have our new AI search feature. So you know Mold is a marketplace. So clients can look for freelancers to help with their projects. And now because of course AI, you can do that with AI so they can just say what they are looking for or what project they have and then AI magically tells them, you know, maybe this is the right person for you to help you with your projects. So I'm seeing AI everywhere, like directly at my work, but also the features that we translate or localize outside with the tools that we collaborate with. So really, really, really exciting to hear that. And of course, we love to understand. And this is why we are here. We love to understand you talk about lean organizations. You talk about lean in house localization teams. And right off the record and we're talking about it, right? What is AI doing to help in that process? And I believe as well, you know, as humans, we don't want to be wasting resources. We don't want to be doing things just because, you know, I feel is enjoyable. It's like it must have some sort of value whether it is artistic, business wise or just personal, but, you know, we have to think about that and talk about that. So tell us about, you know, that leaner in house localization process that you have. Definitely. And I think you say something. You say how does AI help? I think that AI, it does help, but it's also a bit the push that maybe we needed to actually see leaner more strategically placed localization teams where before maybe there was a big mix of what the localization team looked like in a company. Today more and more localization is being positioned in a more strategic position. And I think that's also thanks to AI in that, for example, when I talk about a lean team at mouth, I basically, I mean, leaner than me, I think it's just me. So, you know, that's a good leaner than that. So it's me and then I always say we, so I want to talk about the localization down like we, because lean doesn't mean that you're alone, that you're an island, or that you're, you know, you're a lonely planet, or that you're actually part of this galaxy. And you have to collaborate with all the stakeholders and I always say we because absolutely I am lean in that I am the localization manager, but then that also gives me the possibility to work directly with the people that know what they need, where their challenges are, and what the roadbloats are. So just to give you an example, and this is pretty like we did not plan this any, you know, like, but this happened today. I talk with the people from the marketing team today because they do a lot of manual steps on their flow. And I am now centralizing localization at mouth as much as I can. So I started with legal, I have with some with compliance product was already part of it. I'm looking into marketing now. And they were, I was talking with the people who actually have to do the manual work, you know, the click and they were telling me yes, we create the email. We could be pasted into a document. We run it through a translate system where we are not going to say the name, but like we run it through a place where we translate it, then we get the translations and they could be paste them back into our tool and then we publish them. And I was like, oh my god. And I knew this. I knew this is how they were doing it, but then having someone who actually has to go through that pain every day, tell me that. And you see it on their faces like this is not what this is not what we ambition to when we when we decided to be in market years, it's not copy pasting content into. You know, I know more I'm much in translation engine and then copy pasting it back to just, you know, somewhere else. And then manually pinning the person who has to review it and then manually getting the reply back from them. So it really is. It's the only way in which you can do it. When you have a link localization team, because nowadays with AI, you're expected to have a lean team because you're expected to be able to automate and delegate a lot of your tasks on AI and automation. That also gives you the opportunity to be directly in touch with these people who have these real problems that you can really help. And I see it. I saw it today specifically with marketing, but you see it also, I don't know, with the compliance team two years ago when I joined or with the legal team. And it's, I think it's a bit scary for localization industry folks because we've always been used to having big teams within house translators even or maybe with project-made dedicated project managers, dedicated people to quality, dedicated people for a lot of things. And now it's like one person does it all. And companies are just making their team smaller and smaller and you think, when am I going to become replaceable by the machine or the automation? And on the one hand, I understand that. On the other hand, I think there is still a place for those teams in very big corporations when you need that kind of role. And then there are companies like Mount where there are 700 people working and you have one person doing localization and maybe there it makes sense, you know, because it's a smaller team, it's a more agile environment where things can happen quicker. And there you can work with a lean team. And thank you for making that clarification of like lean versus isolated. And then for some companies, probably localization team looks a lot like people that also take care of marketing activities and then take care of other content responsibilities. And then, you know, it might look like it's a bigger team, but you know, it's, it's a very different function. And one of the things that we were also talking of the record was where this approach is that we see right now on the one hand, you have the orchestration element of it. And on the other side, you have a bit of a one stop one shop single flow type of processes. And it will depend on what your goals are and it would depend on so many other elements. And I think that's the beauty of like where value will be added from from what you've seen and what you are experiencing right now. You've talked about a little bit about how localization seems more of a strategic function now before it used to be a nice to half of foremost. So now it's it's changing that it's changing its own nature becoming more strategic, which probably means it's going to have to get involved or you will get involved in some places in content and decision making. And perhaps that's also happening because of the better understanding of the technologies, but also how artificial intelligence has made everyone rethink different things that they do. How do you see that happen in a mold? How do you see it in the industry, this orchestration versus single flow and then all the day. dependencies that we see around. So I think to clarify, well, not clarify, but I would say maybe localization was very cost center oriented before. And now it's very, it's more something that brings value to the table. And I agree that the mentality has changed a lot, but it's not something new in the sense that we used to have the waterfall, you know, like, I request my project and then that person checks the project and then they send it into translation. And then the translations come back and then, you know, with questions, and then we ask the questions, and we send them back. And today, a lot of translations are happening all the time. So it's continuous translation, it's happening right there. And at Mount, we're doing that, right? So over content as soon as it's available, it's automatically much translated. And then the humans are coming in and they are reviewing it and everything that the machine didn't do well. It's being overwritten by the human review. But of course, that also means that in the future. So that is just one way of doing it. The continuous localization that we have today is just one day of one way of doing it. And it works for some types of content. And maybe I will even say for some language pairs, and it might not work for others. But then there are other ways to do it. And I think it's really funny that you talked about the content creation part and how sometimes it's not about translating, but about creating the content from scratch, maybe even with a prompt, you know, not even by a human, but even from an automation point of view, the quality of the content that you get from an LLM, when you ask them to translate a piece of content, it's not the same as if you tell them, this is my brief. And this is that of article that I want in this language or for this market, and the quality changes a lot. Because, you know, it's just and it's you can see it when you're reading through these articles and you can see the differences. And so of course, also in terms of translation, the more you move upstream, the closer you get to the strategy part of things and to the decision making part of things, the closer you get to when the content is being generated and created. And that's where people start to question, does it make sense that we are actually creating this content in English and then translating it or does it make sense that we are actually generating this type of content in different languages from the start? In terms of the orchestration, I believe we needed for some things. There are some processes that are very complex. And I just as an example for product, where you really need to make sure that there are a lot of moving pieces that are coming together. And every time so I the provider that I use has like an orchestrator tool. And it's really complex to use because you need to be you need to know about engineering in order to use it. But whatever I can, I'm like, can't is there something some connection that I can use out of the books that already exist so that I don't have to go through an orchestrator to do it. And that's always my mentality is like go for the you know, the easiest straight path. But then if I can't do it and then I have to do it through orchestrating the flow, then I have to do it. There is no. You know, there's no way around it. It is true that maybe before I would have needed to be an engineer to do that. And today I don't anymore because with the help of AI and some like blocks, you can play with things and you can do these things yourself. I do believe that the higher up we move in localization, the closer we move also to developers to the topic of internationalization and to the topic of globalization, which is again related directly with the content generation. And it's very interesting, which you mentioned about the intricacies of the different orchestrations that you've probably have to build depending on the type of quality that you want. And I get already tell like if in a flow, you are able to say, take this original article and then create a description of it to be create to be trans created. That's very different than say translate. I can already I can already imagine that in fact, probably it will get much more intricate over time. And those this is my personal intuition, those that navigate that well, the how you orchestrate, how you design are the ones that are going to get the best results. So as this as these functions are evolving, do you think having having these skills gives your company a competitive advantage, gives your localization structure a competitive advantage over those that, you know, perhaps say, no, I'm just going to keep it simple, you know, just going to use this one stream and I'm not going to care about intricacies and I don't care about the final markets or culturalization, which is where I will go with my next question. But the thing is, give some competitive advantage to localization teams to think this way. I do. I think that's the type of people we are in this industry. I think we've seen localization emerge, not going to say born because it was there before before I started in it. But we've seen it emerge and we've seen it more like, you know, evolved and adapt and change and the technologies that we have today are unthinkable. But when I started my degree where you had to like manually allow segments, you know, I told you know, like today, if we told someone of 20 years ago what the possibilities are today, their mind will be blown. So I think it gives it gives that the company a competitive edge. But I also think this is like the minimum company should be expecting from the localization managers that they are able to make use of what's out there, that they are not. It doesn't mean that what's up there, it's, you know, it doesn't mean that what you're using today is not good. It's just evaluating in every moment what's best for you on just to give you another example. When we did the legal translations, the team kept coming, the legal team kept coming to me with very small changes. And they were done in Google Docs, you know, and I didn't see the need to automate to centralize. And then they came with a really big document that they were like, I may have this translated in a Google Doc. And that's when I say, no, wait. Because we didn't have the need before and we didn't have the opportunity to improve this before or the return on investment was not justified. But with this document, it is because the savings that we can have on this by using the right process and streamlining that they now they are here, now they exist. So it's not about always, you know, improving everything immediately. It's about keeping your mind open and deciding, okay, now this is the right moment to improve this part of the process or to improve the process for this type of content or for this language bear or specifically with a yet. So I think it's more that. And you know, as you're looking for this improvements and you're making these decisions, we'll probably have to make a decision about the elephant in the room. And it gets uncomfortable when you start talking about like translators, specifically in this in this case, you know, interpretation is probably a different a very it's a very different conversation. But translation, right now, a lot of it, we probably all agreed a lot of it will be done through machine translation. A lot of it should be done through machine translation. A lot of it was not done at all. And now we'll be done somewhat. And I think that's that's an improvement. And then we start thinking about the other side of the pie, right, which is globalization, internationalization. And then you think, okay, that's the when does content need to be culturally aware, culturally sensitive, when does it bring us a higher return on investment? And is there a place there for translators for linguists for those that care about culture and the meaning of words behind what words are someone is going to have to train the large language model and tell it these word in these five countries is different because of that yet, explaining those sensitivities is going to take us a few decades before so many people have expressed what it feels like for the AI to actually express something like that as well. So yeah, I guess the question is it's out there on what's going to happen with the translators is there a place. And then it's it's culturalization really important for the future that we shouldn't delegate it to AI just yet. I know that's a lot of questions my apologies. Which is okay. And go with it. No, I mean, I'm going to go with the one that's close to my heart. And that's translators because I think and I have a lot of friends who are translators. I was a translator myself at the beginning of my journey. And I talked with them and they're you know, there is anxiety out there. And I understand that and I work at my work that I did with freelancers, so with freelance translators. We don't work through LSPs. We work directly with the person. So that's not to say anything against LSPs, but what I mean is I have a personal relationship with the translators that I work with. And it is scary up there. But it's just I think the role it's about. I think they're still needed, very much needed, because, and I think we've talked about this, but it's just it's not the same. What a machine can produce and what a human can produce is not the same. What a human produces feels a lot closer than what a machine produces to us. And I think the question was, you know, because for multi-ingual, I said, you know, imagine if I wrote an article with AI, you would know. You would be reading it and you would be like, this was generated by a machine. And the same way when you generate your content or your translations to AI for your product, for your brand, the people know that people know a human was not behind it. And then it depends on what type of company you are and what level of closeness you want with your users, not everyone wants to have the same degree of closeness. Far out, it's important because we work directly with freelancers like we have the rule, freelancers rule, or world. So we want to be up close and also for the companies, we are there to support them. So we want a really personal approach and something that when they read it, they know we didn't just put a prompt on a machine and you know, then spit something in there because it was easier or cheaper. So the humans are still involved in the process and all pieces of content that we translate through a much in translation are reviewed by a human. In the future, from a cultural point of view, because it's not only about the culture within the content, it's about how human the content feels, right? And even if you say, okay, well, in the future, the content feels human because we've managed to replicate that humanity. Okay, what about the culture aspect of it? And then there are levels to it. It's not the same to localize a marketing campaign where things can go very wrong and a lot of money was invested into it and just doing it wrong could mean failure. Or you know, like you want people to tell you apart as a company, but for the right reasons. You know, like being deferent shading yourself from others, it's not always good. It's only good if you're differentiating yourself with the right reasons, but if you're differentiating yourself because you're Amazon and you're much translated, you're a product in that I think it was Denmark, some northern country, and then it was much intranslated and there were horrible things there. And it made the news, is that kind of publicity that you want to have? I don't know. I don't know how that impact would say of maybe what's super good for them. But you know, it's and I always say the same thing. The way in which you're treated, the quality of the food that you eat, the price of the food that you eat, and the speed at which you get your food at McDonald's, it's not the same as if you go to a Michelin star restaurant. And there are a hundred of in-between. But when you're a Michelin star restaurant, you buy your ingredients, you cut them by hand, you cook, you know, you have people cooking everything, nothing is frozen, nothing is made by a machine, nothing is, everything is fresh, everything is handmade. So that's the same level of comparison. If you want to be the Michelin star restaurant, then you need a human there. If you want to be McDonald's, which is fine, it's just a different category, then by all means. Machines translate away and never having you in review, it doesn't matter, people are not having great expectations about it anyway. I think that's probably the big point about it. You did mention something before we get, because I would really like to dig a little deeper as well into what's important to translate them all. And what is that there? But before that, let's get a little bit the rest of your team. So you did mention, of course, you are there, but you are also looking for this content to fill human. I've heard that the best business strategies to care. And if you care, you know, you take care of all these things, when you collaborate with your other teams and you know, you're supporting the entire company, what makes a collaboration, a cooperation, successful? What are the elements that you see there in your communication and in the way you approach it that help you accomplish your goals from the localization perspective? I think caring, when you really care not only about your final user, but if you treat your colleagues also, not as a user, but in a way, yes. So if I am talking with, just to give you an example, now we are centralizing marketing, but a year ago, I already brought up this initiative and it didn't go through. So for several reasons, and I thought about it, I was like, okay, what happened, what went wrong? And then I had to repeat the same with legal. And there it went through, it went well, and we managed to centralize and you know, it was a project that I wanted to do on Alchemy. And I thought, what's the difference here? And I think that the difference was, I had not taken the time that I should have taken to really truly understand what was driving the efforts of that team. So I was selling them, not selling them, but like highlighting the things that I thought were important and just to give you an example, maybe quality. And maybe they were thinking more in terms of money, so savings maybe. And although that was part of my conversation, or part of my conversation of my business case, maybe I didn't focus enough on it, and there was a lot of noise around that part that I was trying to get through. And so we didn't go through clean. And that meant that they thought, oh, that's a lot, or maybe in their head, it was a lot, and then they were like, let's think about it later. I was also going on maternity leave. So I thought, oh, you know what, yeah, good. When I'm back from maternity leave, we can talk about it again. In hindsight, every time that I've had a project go through, it was because I took the time to talk with the people to really understand what their problems were, to really understand what they didn't want to be doing, and what they wanted to be doing. And I've had difficult conversations with people who have told me, look, I would like to quit only because of the manual work, you know, because I do so much manual work that it gets to the point when I'm like, what was I hired to copy paste? And I understand that it's not easy to hear that, but when you have that kind of conversation, and then you go back to them with the with with solutions to help them feel better about the work that they do, of course, they're going to want to work with you and don't want to adapt, adopt the measures that you want them to to work with. And so I would say the most important thing is really to understand, to try to understand that we are when we are trying to solve people's problems, we shouldn't be focusing on what's going to be the benefit from us from an localization point of view, or what we believe is best from an localization point of view that really leads them to what's important to them, and maybe it's not your priority. Maybe it's not what you think it's most important, but it's what's most important to them. And then once that door opens, you've got to work on the rest. The other things that you think are also important to you, maybe quality, I don't know, like just to give an example. But yeah, I would say that's that's important. You really get to know the people that you work with. And I love to talk to you about quality. That is one of the big conversations that we also see before we get there. So yes, you work with the teams and what you're saying makes a lot of sense. If you don't care why this is important to them, you don't have that level of empathy. It's going to be really hard to help anyone in the world. And of course, their goals are related to the metrics that are around the things that you do and what the company wants to achieve. Why don't you tell us a little bit about months? Of course, it makes a lot of sense that you have your own freelancers, that you get through your own network. Tell us a little bit about malt. What do you do internally? And what things are important to be communicated in different languages so that we kind of understand, I'm like, where are you, where are you feeding? Yes, so I think, well, the type of content that we translate or that we localize at malt, you have to think that malt because it's a sort of like a marketplace, but there is also a lot of compliance involved. And of course, compliance means that depending on the country where you operate, the laws, the legislations are different. So there are different levels of what you localize and why you localize it. So say the product. The product itself needs to be as general or generic as possible because it needs to fit with a lot of different realities of a lot of different markets with a lot of different needs. And that means that there in terms of the product, there is a big investment on internationalization. So making sure that we create the illusion of localization on the final user. So there we're talking about variables, we're talking about, you know, really modifying the content for the user in a way that it makes it look like it was done specifically for the user when in reality, it's all generic in the back. So that's the illusion of the product. Then you have the help center. And the help center, we have some generic parts, but then we have some very specific parts that are related again to the compliance and the regulations in the different countries. And their translation is impossible. Why? Because I cannot translate the legal requirements for opening your own company as a, you know, independent worker in France for a user in the UK. It's just not going to work because it's not the same. So there it's full translation or not even translation, full creation from scratch of the content because otherwise, it will never resonate. Then we have the website and on the website, we have dedicated domains. So we have multi-f-r, multi-s, multi-k, multi-e. And those websites, the value proposition changes because maybe, just to give an example, maybe the freelancers in Spain, when we tell them you get paid in each day, and we have an insurance that you get paid when you complete a job for a client. Maybe for them it's important because maybe in Spain, there is a problem with freelancers getting paid by companies. But maybe in Germany, that is not a problem. freelancers always get paid. And maybe there we want to highlight the security of working with us through our platform because that's more important for the client user. Whether it's the company or whether it's the freelancer that they are only going to work with people that have been budgeted by our teams. So that have gone through there, checking their documents and the real people with real companies or their real freelancers doing real freelance work. So a lot of that content is generic and translated, but then some strategic parts of the content, for example, the value proposition is localized. That's another part. And then you have the legal part, and for the legal part, of course, the same applies. You have the data policy and the terms of conditions and like that. There are parts of that change during some other parts that are the same independently of the language. But I think, did I like to say a specific type of content that you would like me to talk about and about mouth, but I think that's the most important part. So the website through where the prospective freelancers and clients come in and what's important to them so that they know that with us, they are safe, whether it's security, or whether they work with or save, whether they get paid or not, then they go into the product, this purance is to be more generic. Although for them, it looks personalized, because it's internationalized, and then they help send there where they get the help that they need for the marketing, which they operate, actually. - Thank you for sharing. I think those details are gonna be really good for a lot of people to do this. And whether it's done to the humans, whether it's done through machine translation, independently of that, even if it's just copied from a government website, because that's exactly what you need to place there. There is the whole conversation on quality in the industry happening, and a lot of people are saying, hey, at some point, this conversation is not gonna, it's not even gonna be a question, like who's now following up with website testing or debugging? Like a lot of those things now are done through software and technology. And then of course, there is that extreme camp, probably true for machine translation, there's gonna be plenty of options to just make sure that this very basic type of content gets that level of quality assurance. Now when we talk about this more complex elements that need to be translated, or even just created from scratch, how do you think quality is going to evolve? What are you seeing right now in that conversation around quality? - It's like, I've had a DiJavu, 'cause I had this conversation this morning, I did an interview with a girl who's doing her end of master's thesis, and she's interviewing localization professionals, and I talked with her, and she asked me at the end, like I have a question about quality, and I was like, oh no, quality. And look, I think I'm gonna give you an example. This is very visual, so I need you to make an effort to try to look at this with me. So I have an exercise that I do, when I train people or explain people about quality, and what I do is that I show them a painting. It's a painting by an Italian artist. He's a very good painter. Amazing. The painting is just like a scene of an old dentist removing a tooth. So then I ask the question, do you like this painting? And people normally are like, no, because it's disgusting, you know? And then I'm like, okay, is this painting good? And the answer is yes, the painting is good, because it transmits an emotion. You look at it and you go, ah, ah, ah, the light is amazing. The proportions are perfect. How the artist plays with the light. I was sent to the painting later, so you can see it. But I think if you're following me, you can see it. You can almost feel it. And that's what happens with quality. A lot of people have opinions on what's good and what isn't, but in reality, it's what they like and what they don't like. And then you have quality, that is whether it's good or it's not good. So just, this is just the basis on which I build on my idea of quality. You have three levels of quality. You have the real quality. So are the proportions good? That's it. It's the light, the shadows, and are they correct? And that's what I call the basic level of quality. And that can be do by Emma Chin, because if you set the right criteria, it can tell you actually, you know, there is a dot here and it's not there. You put this terminology in the glossary and they are not respecting it. So, you know, that's just, it's obvious. It's there's nothing subjective about it. Then you have the quality of, yes, but is this what you meant? So does this translate when the user sees it? Do they feel that there has been translation? Is translation invisible? And to me, that's intermediate level. And what we do, or what we're starting to do this year, is that for tier one features, the product managers and the designers, they can ask for in-context review. So that means that for that specific feature, they can ask the translator to go into the product and look at their translations in context and provide feedback. Tier one features why? Because you cannot justify doing these for a feature that nobody ever visits. That is just there because you have to have it there but, you know, how do you justify that effort? And then you have the quality level of when you saw that painting, did you feel what the artist wanted you to feel? Did you feel immersed? Do you know, it's more about, it's not even about whether it's a good translation, it's about whether it's better than good. And for that, I think at the end, it's about the people that sell your product. How do they perceive the quality of your product? Do they walk into a meeting room to sell your product to a user when they are going to do a client and they are going to do a demo? And are they super comfortable showing your product? Or are they like, oh, God, these sunsins, I always have to explain it. Or there is a typo that's been there for three months and I hope they don't see it. And they're trying to sell, to tell the users, we are the right company for you. And behind in their mind, they are not convinced that the quality of the content in our product is amazing. And I think you could feel that sometimes, when you're buying something, you can feel it when the person selling it to you, it's not 100% convinced. So what we want to do at the end of this year is to sit with or front-facing teams. So with the people that actually have to go out there and sell our product and ask them, what makes you uneasy? What makes you uncertain? What makes you feel like, oh, my God, please, let them not see this? Or oh, my God, no, again, I have to explain that part of the product because it's not clear enough. It's not intuitive enough. That's it for them. So that when they go into a sales meeting, when they go to a client and they show our product, they are like, yes, this is like the best product you're ever gonna see. And maybe they feel that way, but not about the content. So the content on our product, just put it, the content in our product is amazing. You're not gonna see better content anywhere else. - This is very interesting because I've recently been looking at websites where their business is content or consultation. And then one of the things that you realize, specifically for English is that this is not re-vergitated content. This is well thought of content. This is content that they've spent time working on. probably if they go to a another language, they are going to just create instead of like transcript. They might be some trans creation, but there will be some like effort put in there and you're saying, hey, this is kind of like a visceral feeling. And it ends up being part of like where leadership wants to go. And you know, this is dependent on leadership. And you know, we already talked about this, so we don't have to go in in depth about that again. But you know, if leadership is not thinking about, you know, being globally aware and then growing and then having these these competencies globally, then they won't have them. And they won't be able to compete with the companies that do care about these these sort of elements. And of course, QE has, like you said, this technical element, you can say this is a 10, this is a 9, this is a 7. And then you can have all these feelings around around it, whether it's right or it feels wrong. And probably you could put also a scale there. From from what you were doing, what are what how do you measure measures success? Are there any KPIs any metrics or or is also this part of like the strategic conversation with the leadership? I have a million KPIs. I love measuring things like, um, some for example, I've have KPS that we've all heard. I don't have like any new like secret sauce or anything. But um, just to give you a couple of examples of KPIs that are very important to me. One is number of requests without issues. So it's the number of translation requests that go through. So the person requesting the translation asks for the request, we complete the translation, the translation goes back and nothing went wrong. Um, that for me is very important. When I started at Mount, um, 45% of requests had issues and now we are around 11%, 9%, um, so that's really good because all this, every time you have to ask, it's time effort that you spend there. So it's good to not do that. Um, and you know, that means the processes is smooth. Another thing is time to market in a way, one of the issues that we had is, okay, when I asked for the translations, how long do the translation stake and sometimes if the PMs or the developers have a launch date or they have to perform tests and maybe they need the translations in, not the final translations, but yes, the translations into perform tests. They don't have them. So by introducing the machine translation immediately, so immediately, machine translation on top of you, and translation that meant immediate time to market. So the moment the content is in our, in our system, you have a translation, which means that you could potentially immediately go to market. And that's another KPI, it's a time to market, which now is basically zero. Um, another KPI that is really important to me, um, or like that, I feel, uh, closely to is efficiency. So the amount of time that the developers and product designers or other people spend on requesting a translation, um, that's also going to go to zero, hopefully in the next weeks we're working on a process, we are provider to do everything automatically. It's almost automatic today. So it's got really, um, um, it's got really well over the past, past months, but that's another thing that I want to do. Uh, what other, I have a lot, um, let me see what else. Well, I also have, um, of course, volumes, uh, how much volume we process, how much the volume is going up, quality. So like the, what I said, like the basic level of quality, how that is going in terms of scoring, um, other KPI's or internationalization. So at that 100%. So when I arrived, zero percent of the content was, uh, dynamic. So I wanted all the content to be dynamic, so introduce variables where we had static components. So that was another one. Um, yeah, I can set you a list later if you want to, but those are like some of the ones that are like, uh, and that, that, that's great. And, and probably gives also audience and impression on, you know, what's under the hood and, and, and, you know, why, why it sounds really good because of course you're keeping track of things. It's not, it's not just, it's just not, it's not just, uh, the feelings around, around, around the things and we're coming to an end of our conversation. It's been, it's been great. I hope we can come back to it and talk more about the content. I would love to. The content strategy will probably create some of that content there. I want, I want to, to end on, on a strategic note, um, on, on what you're doing and, you know, been a lean team, but been highly strategically involved, uh, in, in all the different departments at the organization. How, how does it contribute from your perspective to, to this long-term globalization and internationalization, uh, strategy at, uh, your organization? Look, I, it's, um, I will come back in a few months maybe to talk about a project that I'm working on now that I can't talk in detail. Um, but I'm going to use as an example the idea behind it. So, as a localization manager, when you talk with a lot of different people, you have a three six, a little bit of like an outside 360 view of what things are connected together or what things are sort of like weaving through the company and what things are a general thing for the company. Um, and sometimes we think that's not my job as an organization manager to raise that concern and to say, have you thought about this? It doesn't even need to be a new idea. It could be an idea that has been proposed several times and that you saw and you thought, hey, why didn't anyone follow up on this? Because this was a pretty good idea. This could have solved a lot of problems and this is probably direction in which we should be going. Um, so it doesn't need to be yours. It might be something that already exists and that when you see it, you think, huh? And so I think from a strategic point of view, you as a localization manager have so much on your plate, so many fires to boot out that sometimes we look down toward desk and we go, go, go, go, go with our projects and the things that we want to do and achieve. And sometimes this, this project, these bigger things come across and we're like, ah, yeah, but not now. I have a lot of things to do or oh, yeah, but that's huge and that's not my job and somebody else will do it. If it's about globalization, if it's about how your company positions themselves at a global scale, even if it's about internationalization and you think, well, not there yet, it is your job to bring the attention to that. Um, and being connected with all these different departments will help you see those opportunities. Push for those projects or opportunities that you're really believe in and bring value to the company ultimately. So I think when you are very well connected with everyone at a company level, it gives you a vision that nobody else has or very few people in the company have with the mentality of someone that thinks about global, about how do we, how do we make sure that we arrive to as many markets as possible? And I'm not going to say anything more about it, but maybe I will be back to talk about it a few months when it's actually you should be back and we'll definitely pencil it. Uh, and of course we have an amazing audience and they're always telling me, hey, make sure that, you know, we, we, we, we ask this different questions that we stay a little longer than just 20 minutes. Uh, so I think we've done a great job this time. Is there anything you'd like to mention to those that are listening right now, maybe your colleagues, your freelancers, anyone in your company, any, any final thoughts? I am well, first of all, thank you. Eddie for inviting me. I'm really happy to be here. I really enjoyed it. Um, second, I am very proud of the process that I've gone through and mount because I'm very proud of the company I work with because I really care. And I think as a localization manager, you also have, if you really care about that, you should really look for companies where you feel that that is a match. Sometimes I hear localization manager say, oh, but I try and I try and I try and I try and it just doesn't happen. And it's like, well, maybe you care so much about it, go and find a company that really really cares about that human connection between your product and what you do for the people that you cater to. At the other way around, if you're a company and you have a localization manager and you really, really care about the final experience and your localization manager is not considering that and they're just like hyper-automating everything without considering the human aspect that maybe change. The person that is leading that department in your company. And just for everyone, every translator out there listening, I don't think that you're replaceable. I think that we need you. We need the human touch now more than ever. If a company wants to distinguish itself from the automated generic view, they need the human touch. And the companies that really care about it, they will still work with you. And that's good. Sometimes you know that we'll filter out the companies that maybe. That's not so great to work with. and for all localization managers out there, remember that AI is not there to do your creative work, so you can do the dishes and clean, but AI is there to do your dishes and clean your floor, so that you can get creative and actually do the fun part of your work. That's it. >> All right, Thresor. Thank you so much for everyone who is listening. This was our conversation with Teresa Torón, huh? From Malt. Remember that here at Localization today, we talk about what is driving innovation in our industry. My name is Errieta and the CEO, Irmaulta Linguel. Media, Teresa, thank you so much for joining us. >> Thank you. >> All right, until next time, everyone, goodbye.