B (3:24)

Yeah, that's really the, that's really the trick, right? So I like to say that good AI, right, is like magic and it's easy to use. Like we all have that experience. Sit down and you start using GPT or, or Perplexity for the first time, you're like, wow, this is easy and it's amazing to use to build that experience is hard, right? That's why these companies are spending hundreds of billions of dollars to build these, these experiences. What we do for our customers in order to make sure that they've got that brand voice, that they've got that high quality is we build each customer a custom trained engine, right? So we take their translation memory, we take their glossary, we take their style guide, tone of voice, et cetera, and we take all of that information and we use it to train these engines. So that way when the initial translation comes out, it's going to sound like that customer. So I'll give you a great example. Both IBM and Apple are customers of ours, right? So they're both tech companies, right? But they have, as you can, as soon as I say this, I think you will recognize that they have very different tone of voice, right? So it used to be in the past that you would rely on the translator to look at a style guide and interpret how that was. Now we're actually building that right into the AI engine so that whether we're translating into French or Japanese or Korean, that that tone of voice comes out appropriately and by the way, appropriately for that country, right? So another nuance is Germans have a different tone of voice than Japanese or French, right? So you can imagine that. So we make sure we build all that in and that's one of the reasons why we're able to build these, why we're able to provide our customers with AI translation that delivers on brand correct tone of voice by country.

A (5:08)

That's awesome. I really like that you highlighted something here. Good AI is like magic and I think you and I both have seen good and bad AI and I agree with that. But what I really like about it is you. Each customer gets their custom trained engine based on their brand, their sentiment. Building that style guide right into it. I think that's amazing. So thank you. And Brian. Working with global brands, for example British Airways, Verizon, MasterCard means ensuring translation and localization at scale with reliability. How do you build a product and operations that guarantee both speed and brand safety when rolling out into dozens of markets simultaneously?

B (5:47)

We're smartling is an enterprise grade technology solution. We, we have thousands of enterprise global brands that rely on us to be part of their process. Right. So 99% of 99% as we talked about, is automated. So uptime is really important. We, we have 99.99% uptime, so we invest heavily in making sure that that redundancy is there. In fact, there's been a number of instances with as you remember, AWS went down, the Northeast went, or the Eastern Seaboard went down. A week or two ago, we remained up. If you remember when CrowdStrike had a problem, we actually, CrowdStrike's actually a customer. We actually help them serve up communications during that period of time. So we work hard to make sure that we've got the uptime that our customers rely on. We also invest heavily in data security and privacy and that's evidenced by our PCI SOC 2 HIPAA high trust certifications. Our customers also value audit trails. So we have a lot of healthcare, a lot of financial services companies, so they require. So we, so audit trails are a big part of what we do.

A (6:49)

Right?

B (6:49)

So we just. Every time you change content, right. There's an audit trail that links back to the prior version of that and all of that's then kind of capped off by a powerful language quality assurance and hallucination detection and mitigation solution on the back end. So as a matter of fact, because of all of that that I just described, we actually guarantee quality, which is unusual in our industry.

A (7:13)

That's awesome. And I'm glad you took a time to really unpack some of that, what you're doing to ensure that reliability, that redundancy. You mentioned automation and uptime, but you also invested in data security auditability. I think that's important. But at the end of the day it's the quality assurance is big with your platform and I really appreciate you that sharing those insights.

B (7:35)

Yeah, at the end of the day that's really key. And there's two ways we offer an AI translation, which is instant and very high quality. We guarantee that quality. If you require an extra step, that's when we'll bring a human into the loop for review. So that's kind of like the build and the suspenders.

A (7:51)

Thank you. And that is important a lot of times for qa. So appreciate that. And the last question of the day, Brian, Looking ahead maybe five, 10 years, how do you envision the translation localization industry evolving with AI? Will the role of human translator change dramatically or what should companies start preparing for now if they want to stay ahead?

B (8:12)

Sure. So we view we're building translation as a service is the way to think about it. So we see translation localization writ large as being an automated centralized utility that's available to all functions within the organization, able to enable once again that custom engine, that high quality on brand translation easily and instantly. So whether I'm in the marketing department, the or I'm a product manager or an engineer, learning and development, legal, customer service, all of those functions will have and by the way, all of their various brownfield tech stack. Right. That you think about. So you've got CMS over marketing, marketing automation solutions like Eloqua and then you've got PIM over in E commerce, you've got CRM systems, et cetera, learning and development applications. All these are brownfield applications that we integrate to natively with one centralized translation feature that has that custom trained engine and language quality assurance that I talked about. So if you can kind of picture that as your in your mind is like this, this engine that you kind of plug, plug in to the Ethernet port, so to speak of this engine enabling that translation of service, that's really where we are very rapidly evolving to. And so organizations, if I'm a CIO or a CTO of an organization, I should be thinking about that. That's not something that's been available in the past. That's a new AI solution and what we see a lot of right now and I talk to a lot of C suites around the world and everyone's under pressure to not only implement AI, but to show results from AI. And so translation is one of those places with this type of solution where once again you plug it in and not only have you plugged in an AI translation solution, but you're able to almost instantly show significant organizational results. So once again, kind of going back to they're able to translate our average customer as a result of implementing is able to translate 3x more content for 60% less cost, 6x faster. Right? That's like a slam dunk in the world of, you know, you put that PowerPoint up, that slide up and talk about that deliverable in your organization. That's a big deal.

A (10:30)

That's awesome. Really appreciate that. And having that automated centralized utility for all areas of the enterprise. You talked, you broke that out a little bit like marketing department as an example. But again, easily and quickly deployable, custom train engine, plug and play ready to go. Shoring results, not hype and about six times faster. So I think that's really amazing. Brian, it was such a pleasure having you on today and I look forward to speaking with you real soon.

B (10:55)

Great. Brian, I really appreciate talking with you.

A (10:57)

Bye for now.