#291 Unpacking the RWS–Acolad Deal

14 Aug 2026 · 33 min · 16 chapters

Ask about this episode

Ask anything about it. ChatGPT or Claude reads this page and answers with the times it was said.

Connect VO and ask about every podcast you hear, including the moments you saved. Add to ChatGPT · Add to Claude

In short

The episode discusses consolidation and AI-driven change in language services. Main topic: RWS acquiring Acolad (Auclid/Acolad) in a deal still subject to regulatory steps in France. Guests/hosts: Florian (host) and Esther (host).

Key claims

AI is “blowing open” the market with many new competitors; RWS will rise to the top tier (alongside TransPerfect) via Acolad’s ~1,200 employees in ~22 countries; the valuation (~2x adjusted EBITDA, about £20–22m) is low due to Acolad’s historically leveraged, PE “buy-and-build” debt structure, later stabilized by Acolad management.

Notable examples

other M&A (T-Works buying Swiss Global; Alphatrad buying Lexic; Alpha CRC buying Pure Fluent; Magna merging with Nageli; RNA buying Viva; Contents acquiring Balio). EU AI Act Article 50: Anthropic adds machine-readable invisible watermarks to Claude output, but translation complicates detection. Other segments: Zoo Digital financials; international organizations hiring AI-enabled language roles; VoiceAI funding (Fish Audio, Smallest AI, Amelia, Gradium).

Written by AI. May contain mistakes. Listen to the episode to check what was said.

Chapters

Tap a time to open that second in VO

AI's Impact on the Market

0:00 to 0:19

Explore how AI is reshaping competition in the market.

“The market itself is being blown wide open by AI, so you couldn't say like, okay, they're now dominating like across Europe.”

Upcoming SlaterCon San Francisco

0:41 to 3:16

Details about the upcoming SlaterCon events and speaker lineup.

“I mean, I've not been to Hyde Park recently, but there's certainly a lot of, I mean, it's like drought land.”

RWS Acquisition of Acolad

3:16 to 5:21

Initial reactions and implications of RWS acquiring Acolad.

“I think hearing you rattle off all the names and companies, I'm like, oh yeah, got a solid program.”

The Financials Behind the Acquisition

5:21 to 8:04

A deep dive into the financial aspects and valuation of the deal.

“So maybe that's why Ben Faz, who is French, felt comfortable doing this because there's a lot of stuff that you need to kind of still do in France to close that one.”

Strategic Implications of the Deal

8:04 to 9:23

Discussing the strategic advantages gained from the acquisition.

“And for Oculus, it was a unique opportunity to get hundreds of big enterprise accounts by means of M &A instead of having to invest in hundreds of salespeople, that would be a lot harder to get some of these accounts.”

M&A Activity in Language Services

9:23 to 10:40

Overview of recent mergers and acquisitions in the language services industry.

“So it's going to be a lot of kind of M &A execution risk and challenges ahead for RWS there.”

Trivia and Name Origins

10:40 to 14:00

A light-hearted discussion on name pronunciations and origins.

“There was some smaller deals also happening.”

M&A News Roundup

14:00 to 15:29

Learn about recent mergers and acquisitions in the localization industry.

“So some Nagali left for Portland and started the deposition and trial firm and then, you know, sold it to Magna.”

EU AI Act and Watermarking

15:30 to 20:20

Discuss the implications of the EU AI Act on AI-generated content and watermarking.

“Now, hard switch to EU AI Act, Article 50, and how it triggered Anthropic to add invisible watermarks to CLOT.”

Zoo Digital Financials Overview

20:21 to 23:08

Overview of Zoo Digital's financial performance and operational changes.

“I'm talking about the point of the act, like that they're doing it because of the act, sure, but like what's the point of the act?”
Show all 16 chapters

AI-Enabled Language Roles in Organizations

23:09 to 28:00

Explore trends in international organizations hiring for AI language roles.

“Now let's try to go through this next topic that I want to talk about, which is a bit of a trend we noticed and we started covering like international organizations hiring AI enabled language roles.”

Salary Insights in Localization Jobs

28:00 to 28:31

Discussing salary expectations and job roles in localization, particularly at OpenAI.

“You're going to have to commute into Manhattan.”

The Role of AI in Localization

28:31 to 29:18

Exploring the evolving role of AI in localization and the importance of human oversight.

“New York City and San Francisco, very different ends of the spectrum.”

Understanding the NYC Campaign Finance Board

29:18 to 30:05

Delving into the functions and responsibilities of the NYC Campaign Finance Board regarding multilingual voter information.

“What does the New York City, let's ask the AI.”

Recent Developments in VoiceAI Funding

30:05 to 31:38

Highlighting recent funding rounds in the VoiceAI sector and the competitive landscape.

“an exhaustive list, but we had three funding stories hit in the past few weeks.”

Rising Competitors in the AI Market

31:38 to 32:42

Discussing established companies raising significant funds in the AI space and the implications for competition.

“Now they're an AI company, they weren't an AI company 20 years ago.”
Hear the part that matters, and keep it.Open this episode in VO. Double tap your headphones to save a moment as you listen.
Get VO free

Transcript

Automatic transcript. May contain errors.

0:00Florian:The market itself is being blown wide open by AI, so you couldn't say like, okay, they're now dominating like across Europe. No, no, no, no. I mean, there's so many more competitors coming into the market generally from all sides, from the tech side, from the AI side.

0:19Florian:Welcome back to SlitterPod. Hello there, Esther. Hey Florian, welcome back to YouTube. Yes, we're back with a podcast from, you know, what's now... It's an arid landscape. Zurich has turned into some Mediterranean landscape, maybe a little bit Sicily style, very dry, very arid. So yeah, welcome. Same in green and pleasant England. It's not so green anymore. I saw some pictures from Hyde Park. Were they fake? Were they AI? Probably not. I mean, I've not been to Hyde Park recently, but there's certainly a lot of, I mean, it's like drought land. Well, there's no drought in Newsland. That's why we need to catch up here on the podcast.

1:01Florian:But first, really excited to have SlaterCon San Francisco come up. So we'll start with the networking and welcome reception on September 2nd. Looking forward to hearing from Ebru Ildirim, who is the CEO of Olang, and she's going to address our participants there at the welcome reception on September 2nd. And then we'll start with the full day conference program on September 3rd, starting at 9.30 a.m. with, I'll share a few thoughts around how we see the industry. Then we're heading over to a panel that I'll be moderating with Teresa Marshall and Semi Altinai from Phrase and Salesforce respectively.

1:41Florian:Then we're heading to a panel with Crisp, our new friends from Crisp around VoiceAI that's moderated by Alex. Then we have Scale AI's Emily Shueh talk about enterprise AI and data for AI. We have Marco Trombetti in the afternoon starting to speak about limits of today's AI translation. Marco Trombetti, of course, from Translated. Then let's go further, an amazing investor panel that you're going to be moderating, like where AI investors are betting. We have Koshla Ventures and Dell Technologies Capital there that, yeah, you'll be moderating that panel. Then we have XTM, a panel on AI localization risk and compliance that Silvia is going to be moderating from our side.

2:28Florian:Then David Lee, CEO of IUNA, the world's biggest media localizer, gets to speak in the afternoon. And then we have Gabriel Fairman from Works and Eliza from Coca-Cola talking about scaling AI across global ops, language AI across global ops. Then one presentation I'm particularly looking forward to, Marlon Duarte from UCLA Health on language access in healthcare. Very, very interesting. And then we got Spence returning, Spence Green from Lilt returning as a speaker. And then you'll wrap up the conference and then we head over to the drinks and the networking reception at the phenomenal San Reaches.

3:11Florian:So that took me, I don't know, two minutes to go through. It's a busy program. I think hearing you rattle off all the names and companies, I'm like, oh yeah, got a solid program. It's going to be a good day. Pretty solid program and starting to look like good turnout also. San Francisco now quite competitive in terms of conferencing, a lot of stuff going on, of course, in that part of the world, especially when it comes to AI, the heart and soul of AI, if AI has a soul, it'll be found in San Francisco. All right. So yeah, you want to join that one if you're on the West Coast, if you're in the US or if you can get your ESTA visa approved.

3:54Florian:So what did you think when you saw that RWS is going to buy Auclid, ESTA? Your two, three second hot take at the time, not after considering the whole thing, but like initially. Well, I thought, oh, did we predict in January 2026 that there might be a consolidation or acquisition of one of the super agencies? So I think we did. It's a perennial prediction. It's a perennial prediction. We don't need to talk about the years that we predicted that it hasn't happened. But yeah, I'm pretty sure that at the beginning of 2026, we said there will be a super agency sale or acquisition. So, but aside from that, you know, interesting.

4:36I mean, I kind of immediately thought about people that I know who might have left RWS to work for Aklad and how they might look upon that acquisition. Welcome back. Welcome back. Yeah. But OK, all of that aside, you know, it's it's it's exciting. I think when you get a deal of this magnitude in our industry and it involves a lot of people and it would have been a very, very complex transaction.

5:04Florian:It's one of those transactions that only a handful of companies in the space would have been able to pull off. And I think it's actually not yet fully closed. I guess it's contractually maybe not fully closed. I'm sure it's going to go ahead, but there's like a bunch of regulatory hoops in, Grum roll, France that they have to jump through. So maybe that's why Ben Faz, who is French, felt comfortable doing this because there's a lot of stuff that you need to kind of still do in France to close that one. All right. Look, the highlights here would be that I think this is going to put RWS kind of at the very top of the LSI food chain there alongside TransPerfect.

5:43Florian:I think Auckland has about 1 ,200 employees across about 22 countries that are going to join RWS. So, RWS is going to be pretty big in terms of headcount after this and revenue, of course. Now, what really turned heads was the apparent valuation, which was like two times adjusted EBITDA. So, basically RWS is paying about 20 million pounds, rounded down. I think it was 22.4, but we're not going to kind of go into the second, like anyway. Hundreds of thousands. Hundreds of thousands here. So basically, roughly 20 million pounds they paid, which is like super, super, super, super low, but the low valuation is largely explained as I also wrote in the newsletter last week by Oculus heavily leveraged capitalist structure and kind of historically debt funded acquisition strategy, right?

6:36Florian:So basically, PE owned, buy and build. So, they spent years kind of borrow, acquire, refinance, acquire, which eventually kind of led to the debt exceeding what kind of the operating earnings could comfortably support. So, the bondholders wanted to get something for that. So, they ended up with, they reclaimed some of it, but not all of it. I mean, go and look up all the French kind of regulatory filings for more details there. But to be fair though, I mean, Aucolet's current management inherited that burden and stabilized the business, headed the transaction. That's no mean feat. I mean, Bertrand, Stalder and his team, they came in, I think in 2023 when essentially the financial structure was as it was at sale.

7:30Florian:Like it was a heavily leveraged structure, 300 million in debt and well, they It came in, stabilized the company, put it back on track and now it actually became sellable to RWS, right? So, yeah, I mean, in terms of the kind of strategic logic, this does make sense. You know, the market itself is being blown wide open by AI, so you couldn't say like, okay, they're now dominating like across Europe. No, no, no, no. I mean, there's so many more competitors coming into the market generally from all sides, from the tech side, from the AI side that this does make sense. And for Oculus, it was a unique opportunity to get hundreds of big enterprise accounts by means of M &A instead of having to invest in hundreds of salespeople, that would be a lot harder to get some of these accounts.

8:26Florian:So yeah, it does make sense. It expands in regulatory or it's another expansion for RWS in regulatory. It's a further extension in geography, I think especially France and Germany and they're going to expand quite well there and there's some stuff coming in from the US. If you look back, didn't... I mean, so Aculad, then it was Amplexor and Amplexor originally acquired some... I'm blanking the name, was it Sagen that they acquired in the US? That's a long time ago. Yeah, that's kind of before my time, but definitely, yeah, Amplex or LiveWire. So it's about 10 years ago and so they have a US presence there as well.

9:12Florian:Yeah, I mean, so execution time now. So you got this, you're going to integrate. Ben was pretty adamant that he wants to integrate fast and doesn't want to run a company that's essentially 25 separate companies. So it's going to be a lot of kind of M &A execution risk and challenges ahead for RWS there. Cool. Yeah, I mean, congrats to RWS, big move. Shares went up. Let me check now live as we're recording this. It was like 15 % on the day it was announced, I think, roughly. Yeah, so still up, still up even a little bit more. So year to date, let me pull this up here. So year to date, RWS shares are up 38 % on six months, they're up 50%.

10:02Florian:On last month, they're up 54%. So basically, yeah, investors seem to be regaining confidence in that RWS is going to end up as a winner in this game. I mean, it's still a far cry from like the 2022 all-time highs, which was like... That was the heyday of the kind of... They do a lot in IP and yeah, pre-AI. But definitely off the lows. I think the lows, I mean, maybe like up 70 % or 80 % from the lows. So yeah. Cool. So there was more M &A. Why don't you walk us through this? There was some smaller deals also happening. Exactly. We can't forget, even though there was that big breaking news story.

10:48Also other M &A or acquisitions happening in the language AI and solutions industry. So we had most recently, we had the announcement that T-Works Group acquired Swiss Global. So, yeah.

11:05Florian:Congrats to Matthias. Matthias, yeah, I think they're quite familiar. Matthias Trimpe, he was my CFO back at CLS. Ah, there you go. And then he set out to start Swiss Global. So apparently he'll be leaving Swiss Global, but the other co-founder, Mark Cheetham, who is COO, will lead the Swiss business as managing director of Switzerland. So that's that one. We also had news from Spain where Lexic Language Solutions has been acquired by Alphatrad. Lexic being a specialist in medical and pharma translations. And the Alphatrad CEO, very bullish on being able to double company's revenues in a few years.

11:51Then we had a couple, well, I think one in the UK here. We had Alpha CRC acquiring Pure Fluent. So that one was completed in June, but I think only announced slightly later, focused on technology, e-commerce and life sciences. In the US, this is sort of probably quite a large transaction. This is US legal sector. We had Magna Legal Services merging with Nigeli, I think that's how you say it, deposition and trial. so very much legal sort of legal outsourced services focused but they have Magna that is who we've kind of been tracking for a while because they've been doing quite a lot of acquisitions including those focused around sort of interpreting in particular so they have what was it so Nageli brings with it 140 employees so not not tiny acquisition and now the total workforce of Magna is over 1 ,100 employees, including 2 ,000 interpreters.

12:59So, no small interpreting operation that they have there.

13:03Florian:I'd like to squeeze in a trivia. So, that name, I wasn't familiar with that. How would you pronounce that? I think I said it wrong, but Nageli, maybe? Okay, no, no, no. So, let me try to give you my kind of etymology of that word. So this is 99 % this is a Swiss name and it will be with the two dots. It's like my name, like A-E is A. So it's Nageli, which is a Swiss name, which literally means small nail. And yeah. Small nail deposition and trial. Okay. So small nail. And it's one of the most famous graffiti artists in Zurich. So there is, I mean, that name, like Nageli. he's a guy he wrote little kind of stick men or what do you call them stick figure stick people so he started graffiting stick people in Zurich I think in the 70s or 80s and nobody knew who he was for like three decades he was like the Banksy of Switzerland yes you're right exactly and I think eventually people found out who he was Anyway, that was a quick de-itre into the name Nagali.

14:18Florian:So some Nagali left for Portland and started the deposition and trial firm and then, you know, sold it to Magna. Probably not the same person. Not the same person, but it's a marginally kind of common name. So maybe there's some family relations to the stick person, graffiti guy. Anyway, move on. Yes. Well, so we had another acquisition of Lisbon headquartered Viva translations being acquired by RNA. So that's the expansion of a Brazil based LSI into Europe. And then we had news of contents.com or contents, I think they're now called acquiring Balio, a FinTech platform based in Barcelona. Contents is like a little bit on the periphery, I suppose, of the core localization AI.

15:07industry, but we did have them talk at SlaterCon a couple of years ago, and they have in their history acquired translation expertise, capabilities, and they're on a bit of a mission to do a buy and build acquisition strategy, becoming a language AI and orchestration platform headquartered in Milan. So that's sort of a bit of a roundup of maybe six weeks or so of M &A news.

15:34Florian:Very good. Now, hard switch to EU AI Act, Article 50, and how it triggered Anthropic to add invisible watermarks to CLOT. Where to start? Where to begin? I mean, A, I have no idea how you add machine readable invisible watermarks to text. So, okay, you go to claw, it generates text and there's probably, I don't know, at this point, I don't know, a substantial fraction of the internet is probably claw generated text. So I was trying to find it out, but I can't even begin to like explain or T-L-E-R how you add invisible frameworks. Wouldn't it just be like, I mean, not that I know, wouldn't it just be like another kind of feature as in you don't see the font, you know, you don't explicitly know what font it is, what font size it is.

16:31So it's just another kind of thing.

16:32Florian:I think it's the actual text. I think it's the actual textual logic. That's how I, what I read in my like, you know, 10 minutes on X. I think it's not like font or anything like that. It's literally the actual wording and letters. So it probably needs a specific amount of text to be able to watermark it. Oh sure. Yeah, like you couldn't just have a one word, oh, that's been generated by Claude. You're right. You may need like a paragraph or something like that. But okay, so let's leave our listeners to go and find out how anthropic machine readable, like adds machine readable invisible watermark to text.

17:11Florian:Now, this apparently also applies to translations as well as other generated text. It's as a, I mean, it was driven by that EUAI Act. And I think initially they wanted to do it, like they're releasing it on models in the EU, but they want to extend it globally. And then it's on all their platforms, Cloud, Code, Cowork, Tag, blah, blah, blah, plus all of the other platforms they're publishing it. So it's designed to survive copy, paste, and some editing. And the problem is... What if you remove formatting? If you're, yeah, exactly. So translation creates a major problem, right? The watermark indicates cloud involvement, but not necessarily AI authorship.

18:08Florian:And as we said before, like short text may provide an insufficient signal for reliable detection. I don't look, again, the translation is a, you can get rid of the watermark by translating, but if you use CLAW to translate, there will be a watermark generated as far as I understand. Yeah. If anybody has a different piece of like information, please correct this, add it to the, I don't know, add a comment on YouTube. The underlying goal, at least of the AI Act is transparency, right? So it's kind of, we want to know if AI is involved in this. But that's going to look so quaint in like 20 years.

18:47Because I was kind of coming at it from a more like a, oh, maybe it's plagiarism, you know, you want to kind of avoid in universities or what have you, like plagiarism, because this would potentially be a way to do it. Or plagiarism or plain copying from AI.

19:05Florian:AI generated content can be detected, but like in five years, 99 % of the content is going to be AI generated. So what's the point? And then if you have a piece of content that's 95 % AI generated and then 5%, you're adding some additional value to it, like why would it need to be, and let's say you generate a piece of content, you read through it, you agree because you're an expert in a specific field, you add 5 % of the text is going to be something you're writing originally, like it's a human generated piece of content. Like why would it be relevant that the remaining 95 % are AI generated?

19:47Florian:Like I don't see the point. And then again, if you translate it using AI, what is it then? AI-generated content? Apparently, it's supposed to be, but like, what's the point? Like, so now you need to flag every AI-generated piece of translation? Like, what's the point? There's no point. I suppose the point is they have, well, they've kind of, they've now aligned with the EU Act, so there is the capability to detect AI. I'm talking about the act. I'm not talking about them doing it. I'm talking about the point of the act, like that they're doing it because of the act, sure, but like what's the point of the act?

20:28You go ask the EU lawmakers.

20:30Florian:I know. And like the EU is becoming this like meme now where like in the US and China, like they're developing frontier models and the EU kind of mandates that little kind of that the cap, the bottle cap is not falling off, right? It's like, okay, yeah, great, you regulate it. Like, why don't we, like, you know, generate? All right, moving on before we get too political. No EU bashing, yeah. Here's a good segue. A company that is not based in the EU or at least no longer anymore in the past eight or nine years is Zoo Digital, so media localizer. Revenue down, so let me rattle through their full year, 2026 financials, revenue down 15 % to about$40 million, mainly on weaker dubbing demand.

21:20Florian:Now, the full year 2025 comparison was boosted by the post Hollywood strike backlog, localization revenues down quite substantially from 30 million to 24 million, dubbing revenue down to 7 million, subtitling up to 1.3 million. Gross margin improved, adjusted EBITDA up, EBITDA margin up, net loss narrowed. That's good because you don't want to have a net loss because eventually you're going to run out of money, which I think they had some issues with liquidity. Headcount down 70%, so they had to let go of people like roughly 400 people now. More fulfillment shifting to India and Chennai as the preferred expansion location.

22:03Florian:What else? They have this fast track service now that's now accounting for 10 % of the revenue, subtitles in three hours, dubbing in 24 hours, and AI is increasingly embedded in their workflows, which I know that Stuart Green, the CEO, he's very conscious of deploying AI in media localization. If you can go back to a podcast we recorded maybe 18 months ago. So he's not like jumping headlong into AI. He wants to have a very deliberate kind of process there. But they do a lot of R &D, I think, don't they? I mean, it's not that they're kind of just wary, they're actively investigating quite a lot.

22:44Florian:Yeah, and so for 27, the full year, they promise or they give an outlook that they return to revenue growth and further profit improvement, but they gave no formal guidance and shares are trading like it's still super, super low, but like off the kind of major lows there. So yeah, let's see. Thanks for all the transparencies. You're one of the very few publicly listed LSI's left. So thanks for that. Now let's try to go through this next topic that I want to talk about, which is a bit of a trend we noticed and we started covering like international organizations hiring AI enabled language roles.

23:27Florian:This is not going to be easy to summarize, so pardon to all of our viewers and listeners, But basically we had NATO, ICC, WIPO, Interpol, FIFA and the IMF all hiring across language services and the common thread there was that really AI and translation tech is now embedded in kind of the core job requirements. So initially we looked at NATO hiring a deputy head of digital transformation, like deputy head slash digital transformation lead, translation PM and computational linguists. That's a one role? No. Well, that's three roles. No, no, no. So that's it. Yeah, you're right. Sorry. That's the law.

24:08Okay.

24:09Florian:No, they hired a deputy head for NATO linguistic services. Okay. And this individual will be tasked to run an NNI-enabled operation. And then the computational linguists is in charge of managing linguistic data, AI translation systems, et cetera. and you need to know English and French. Yeah, and then over to the International Criminal Court that is hiring a head of Language Services plus three translators and they also need explicit knowledge of translation tech and AI tools and then WIPO is hiring a head of the Spanish translation section, similar AI translation, LLMs, TMS, all these good things.

24:54Florian:And then Interpol is hiring an English language department head, also responsible for AI strategy, government deployment. And then FIFA is hiring a Spanish translator that's based about maybe a kilometer from where I'm recording this podcast from, just up the hill basically, FIFA headquarters. Same, AI translation, familiarity, term extraction, quality estimation, et cetera. And then the IMF is hiring and that's an interesting one. They're hiring a chief translator, deputy division chief, but then the IMF head went on record saying like, hey, we're using AI and as an example, they're saying that their translator and interpreter staff fell from roughly 200 to 50 among AI, amid AI expansion.

25:40Florian:So, they slashed their kind of language headcount by, I don't know, 75%, but now they're hiring a chief translator. So, they can pay the chief translator. Yeah, so all of these jobs postings we covered in the last four to five weeks on the website, but then you wanted to talk about a couple of green shoots in hiring, so kind of outliers. There was something in New York and then something in the Bay Area. My one thought about all of the international organizations hiring is maybe they just copied each other in a sense. You know, like we saw this, didn't we have, didn't we do a bit of a like comedy sketch about one tech company basically like totally ripping off the job ad of another tech company?

Read the full transcript

26:26Florian:Wasn't that, that was perplexity off of OpenAI or something like that, where it was like word for word. Okay. Yeah. I mean, the point being, you know, when you're writing a job ad as HR or what have you, you're probably looking at what your competitors or peers are doing and recruiting for as well as what's needed in your own organization. Yeah, green shoots, I don't know. That was more just, personally, I think anecdotally on LinkedIn, I've definitely seen a lot more LinkedIn posts about hires in localization and language coming up on my feed than I have done in recent months. So that's in London.

27:00I mean, that's in the US. It's in Europe. You know, who knows? It could be just replacing headcount that was made redundant sort of a year ago or a year and a half ago, what have you. but I mean yeah we're still seeing I mean I picked up on two of the stories that Slate covered recently and this is really is a bit of a counterpart to the the whole AI narrative that you were talking about in the institutional organizations there's a language service coordinator role in the New York City campaign finance board hiring a language service coordinator in Manhattan. The role sits in public affairs and supports the language access unit.

27:43I mean, that one I scoured the actual job ad and there's absolutely no mention of AI whatsoever. So it reads a lot more like a sort of fairly old school or traditional language role or coordinator role. So you've got things like maintaining TMs, glossaries, style guides. I mean, all still needed and useful, but absolutely no mention of AI within that. It pays around 80K. You're going to have to commute into Manhattan. Apparently, I guess. I don't know how many days.

28:12Florian:From far out of Manhattan. Yeah. And then again, I mean, on the other side of the salary spectrum and probably AI adoption spectrum, there was news that OpenAI is still continuing to hire within its localization, its new-ish localization team. um they i think as we just alluded to they hired their first localization manager about a year ago uh based in san francisco and now they've continued to hire within the department um with the role of course assisted to uh it says i think pilot ai assisted translation workflow so obviously needing experience with ai and machine learning translation technology um but again i I mean, the takeaway, as we said at the time, is that, you know, you've got this AI leader, like arguably the leader, still hiring humans and localization managers to oversee the operations.

29:11Florian:New York City and San Francisco, very different ends of the spectrum. I actually don't know what the, hang on, the New York City Campaign Finance Board, what does this do? What does the New York City, let's ask the AI. What does the New York City Finance Campaign Board? Well, they're supposed to be delivering multilingual election information to voters, candidates, and campaign contributors. At least that role. Okay, let me add to that. The New York City Campaign Finance Board is an independent nonpartisan city agency that administers the city's public financing system, enforces campaign finance laws and promotes voter education.

29:56All right. Well, in the city as New York, you need more than one language.

30:04Florian:Closing off on VoiceAI, which is super active, continues to be, and by no means is this an an exhaustive list, but we had three funding stories hit in the past few weeks. One is a company called Fish Audio. They raised a$52 million, what they call seed round, which is kind of these terms now become like meaningless, a$52 million seed round. That's just one year after founding. The text-to-speech supports 80 languages, voice cloning in 30 from a 10-second sample. What about a two-second sample? I don't know. It gets shorter and shorter. So they're saying they have a$30 million pipeline and 300 qualified demos within 24 hours.

30:49Florian:I mean, note the terms. So if you're going out with a pipeline, I mean, a$30 million, that's a pipeline. I mean, come on. Yeah, sales pipeline. It's a sales pipeline. Who doesn't have a$30 million sales pipeline? Anyway, and then there's another company called Smallest AI, cool name. a more reasonably sized 13 million Series A. They already have 60 people on staff, do the same TTS, STT, speech to speech, etc. Another company interesting called Amelia, which is a 20-year-old company and yeah, just to make matters worse, they're saying they raised a 67 million Series B. So you're a 20-year-old company and you're raising a Series B and they're saying that there's a...

31:38Now they're an AI company, they weren't an AI company 20 years ago.

31:42Florian:Everybody is a serious beef. Everybody, and they have 60 million ARR, which if you actually do have 60 million ARR, you probably would raise, if you're an AI company and you're fast growing and you hit 60 million ARR, you probably raise like a$10 billion anyway. All right. A more, it's actually not more reasonable, but it's a company we know very well, Gradium, because we had them at Silicon London, they added another $100 million to their seed funding and I think their investors include NVIDIA, again, STT, TTS, voice cloning, et cetera. And this market's getting very crowded now. I mean, just to go through some of the more recent funding rounds that we looked at, I mean, Eleven Labs, of course, Cartesia, Gradium, Rhyme, Bland, Valens, Hume, Coval, yeah, lots of companies in that space.

32:33Florian:We'll talk about it, of course, with Chris at SlaterCon San Francisco. Go get the ticket. Looking forward to seeing you there.

From the publisher

Florian and Esther discuss the language industry news of the past few weeks, starting with RWS’s proposed acquisition of Acolad and what the deal means for consolidation among the largest language solutions integrators (LSIs). 

Esther rounds up further M&A activity, including t’works acquiring SwissGlobal, Alfatrad buying Lexic Language Solutions, Alpha CRC acquiring PureFluent, Magna Legal Services merging with Naegeli Deposition and Trial, R&A Translators buying Viva Translations, and Contents acquiring Balio.

Esther and Florian examine Anthropic’s introduction of invisible watermarks for Claude-generated text in response to the EU AI Act. Florian questions how meaningful AI-content labeling will remain as AI becomes embedded in content creation and translation, while Esther points to transparency as the regulation’s underlying objective.

Florian reviews ZOO Digital’s declining revenue, improving profitability, growing use of AI, and shift toward faster localization services and more fulfillment in India.

The duo talks about how AI skills are increasingly appearing in language roles at organizations including NATO, the ICC, WIPO, Interpol, FIFA, and the IMF. Esther contrasts this with a more traditional language-access role in New York and OpenAI’s continued hiring of localization specialists to oversee AI-assisted workflows.

Finally, Florian highlights continued investment in voice AI, with funding rounds for Fish Audio, Smallest AI, Omilia, and Gradium underscoring how crowded the speech technology market has become.

More from SlatorPod

All 39 episodes
#291 Unpacking the RWS–Acolad DealSlatorPod · 33 min
Listen in VO