Can Europe build AI champions?

0:00Funding has nearly doubled to$8.9 billion this year for AI native companies. and at some point the boom times are going to stop being so boomy. When that happens remains to be seen, but some VCs are definitely urging their portfolio companies to go out and fundraise now while times are good. Hello and welcome to the Sifted podcast. I'm Defny, co-author of Sifted's AI and Deep Tech newsletter, covering for Amy while she's away hiking in Taiwan. Today, I'm joined by my colleague and co-author, senior reporter Kai-Nicole Schwartz, and Sifted's head of research, Jonathan Sinclair, who's just published the first annual Sifted AI 100 ranking and report in partnership with USBC N47.

0:48Gentlemen, hello. Hi, Daphne. Hello. Lovely to be here for what is my first appearance on the Sifted podcast. Lovely. So it goes without saying that here at Sifted, we are tracking every development that has to do with AI at the moment. And we have done for the past 18 months to two years now, and we're not getting bored. The US and increasingly China are leading a frenetic global AI race. They're building data centers at pace, releasing top performing models, and Europe is determined not to be left behind. So we've got new champions emerging across the continent. We've got more cash flowing to AI startups than ever before.

1:25We've got policymakers makers who are making AI their top priority. And this is why it was prime time for us to publish our first AI report ranking the 100 most promising European AI startups in 2025. Johnny, do you want to tell us a bit more about the report? What were your findings? What are the exciting companies that you're seeing out there? What kind of fields are they operating in? Tell us a bit more. Absolutely. So yeah, this is a ranking and a report that we've wanted to do for a while uh obviously a lot of attention on ai right now so um really important for us to to shine a light on this kind of next wave of of startups coming along in in europe i think obviously we write a lot about the likes of lovable synthesia obviously having extraordinary breakout success but you know this report really liked really sort of set out to to look at the the next wave so companies valued under one billion dollars so not yet reaching unicorn status so looking at who are the most promising companies ranking them um from from one to a hundred so we kind of did a very extensive scouting process in in august getting to 734 at that time ai native startups in Europe.

2:46That number is already approaching 900. So company creation is really flying right now, particularly in the AI space, obviously. How we define AI native at Sifted is we've got quite a high bar. I think most companies, certainly in this industry, are obviously deploying AI in some way, even us at Sifted. But in terms of how we define it, it's got to be where ai is the core product the the infrastructure layer or effectively the value proposition so less interested in companies where ai is the background feature you know maybe something that's improving existing product functionality or use for internal optimization uh so really kind of high quality air companies that are that are being built in the last few years um you know, with AI at the core of their business.

3:43So, you know, as I said, we ranked these 100 companies, whittled down our list through a kind of scoring model that led us to get to 175 companies that we then judged on competitive differentiation, mark opportunity, momentum, scored them between our editorial and intelligence teams which which kai's well was was involved with um to whittle those companies down to to a top 100 uh and which we're really happy with i think it's a really high quality list uh you know breathes a lot of you know optimism hopefully into into what is coming within europe uh some of these companies a lot of people might not have heard of yet so So there are a lot of young companies, nearly a quarter of them have been founded in the last 18 months.

4:35So, you know, very easy and very quick, I suppose, or quicker than it was before for these companies to find that traction, you know, raise money quickly. The average funding of the 100 companies is 45 million, which for a very young cohort is obviously impressive. um you know that that shows obviously investors have their eye on on all these companies and they're also quite lean so we also do within our team a ranking on on b2b sass companies which which went out early in the year notice quite a big difference in terms of headcount of these companies so the average headcount on our b2b sass ranking was was 136 whereas for ai it's it's 75 so with ai at the core these these startups are building very lean and efficient businesses uh and that's the kind of company you know we want to encourage uh in europe and those are hopefully going to be some of the global leaders to come in terms of how they as well generate revenue other interesting stuff i mean obviously there's been a lot of chatter about as ever you know which countries are leading um europe's ai kind of race so you know unsurprisingly it's the uk france and germany at the top accounting for for 78 of the ranking you know it's been a breakout year i think for swedish ai maybe surprised to see not as many of those companies on this ranking this time around so i think a lot of those companies um lagora the most recently have already reached unicorn status um so there are only three of those on the ranking and dutch companies and spanish companies were were kind of fourth and fifth for the countries that they were most represented um central eastern europe kind of worryingly low uh we only had one slovenian company on there um you know that's maybe the sort of a representation of those companies you know having to move to the uk or the the us early on in order to establish themselves you know we forget that 11 labs was originally a polish company and have since you know made that leap to the uk and then to the us so um you know for order for those companies to make it they might feel they have to they might have to leave that region and then you know i think daphne for something obviously you're keeping a close eye on uh you know the the keenly contested you know as ever paris versus london debate you know who uh of those two cities is the epicenter of ai in europe um this ranking would suggest or this kind of next wave of companies uh would suggest london is taking the lead so with 33 london companies on on the ranking as opposed to only 15 from from paris but as we know um some of the the unicorns already are from paris you know mistral obviously the big one and poolside soon to be maybe joining that club, the Decacorn club.

7:33So, you know, that will go back and forth, that debate. And it's obviously good to have, you know, two massive cities that are generating all this talent. That was something that I found quite interesting. And Daphne, it would be maybe good to get your thoughts on this too. The top 10, none of those companies are from France. I mean, numbers 11 and 12 of Metalier, both Paris-based, Harmonton AI, which is a defense and dual-use company, and H company, which is developing agentic tools for enterprises. But otherwise, I think five of the top 10 are UK, there's a couple of Germans, there's one Swedish, there's one Spanish.

8:09And a couple of years ago, it did feel like Paris and France was really kind of at the epicenter of like a big portion of the exciting AI stuff that was happening in Europe with the rise of Mistral, with the rise of H Company, Meta developing this huge AI lab. Does that feel like it's changing to you, Daphne? I'd say it's plateauing potentially. As you say, Mistral generated a lot of excitement when it raised the seed round two and a half years ago. Now it's created more excitement as it was raised round upon round, most notably, obviously, 1.7 billion round raised last September. So Mistral, I think, created a lot of excitement at the same time.

8:50There was a lot of chatter about Poolside, another quite high profile company originally born in the US and establishing a base in France. It's now debatable whether Poolside considers itself still to be European or not, but it certainly got the French tech ecosystem very excited. We were seeing more entrepreneurs or entrepreneurs to be coming back from the labs of big US tech giants to Paris to launch their own company. And I think it did create a lot of excitement. French tech in the last year has had a bit of a slower year, a much slower year. I think the data shows it as well. What I'm hearing often from VCs is that there's a bit of a flight to quality.

9:32So the big companies will be gathering the majority of the cash that's out there. Actually, if you look at Sifted's data for the past quarter in France, it shows an uptick in funding. But if you take Mistral's funding roundups with that, it's basically the same or even a downward trend. So there's a big focus on the kind of poster children of French tech. I don't know if it's as easy for other entrepreneurs, especially in the current political context. So yeah, I'd agree. I'd agree with you, Kai. There seems to be a bit of a down after a very enthusiastic period a couple of years ago. So another trend that I noticed from the report that I found interesting was that we're seeing quite a lot of startups that are at the application layer.

10:27I think the majority of them are at the application layer. There are a few infrastructure startups as well, but it seems like application or applied AI startups are taking a lot of the focus and more and more of them are developing in Europe. I'm curious to know what you think. Would you agree were the exciting startups that you saw in Europe at the moment, the both of you, when you were judging which companies should be in the ranking, were they mostly application layer based or are we still seeing that focus on infrastructure that we were seeing perhaps a year ago? Well, yeah, I think certainly from a kind of number perspective on both the number of companies on the application level and also, you know, those that are attracting funding, that would suggest that the application layer is where Europe should be focusing its attention.

11:17I mean, so we had 41 of the 100. We're actually on the vertical application layer to kind of explain quickly what that means. Those are applications of applied AI that are focused on a specific industry, whereas you've got kind of more horizontal AI applications, you know, agentic AI, AI assistant co-pilots, etc. That are more horizontal. So, you know, they can be deployed across industries, custom made for your kind of own workflows. Those are probably the ones that we consider to have, you know, not as deep a moat just because, you know, any update from one of the big models in the US can take away that defensibility overnight.

12:06You know, if I was a kind of horizontal founder myself, that would keep me up. but certainly the kind of vertical applications you know you think about for example legal tech you know finance tools these are you know some of the I think supply chain as well some of the the best applications to date and I think you know the most tangible and maybe impactful at this stage you know I think there's been a lot of chatter about related to to the bubble about whether applications are actually being successful so far. I think there was an MIT report that came out that said that 95 % of applications have not increased productivity yet.

12:48But I suppose where there remains the kind of most interest, still these bigger industries where there is a lot of money to be deployed, they're quite outdated in their processes, particularly kind of financial services. Obviously, we've already seen the wave of fintech companies that have kind of transformed that industry, for example. So, you know, it should be the same for AI. And that makes, I suppose, a little bit easier for investors to get behind some of those ideas. So, you know, again, going back to the numbers of those vertical applications on the ranking, two thirds of them have already raised this year, some of them more than once, you know, the likes of Fixer, for example so you know they can go to market and go go to the table sorry very very quickly in terms of fundraising too and that kind of keeps that momentum whereas you know for some of the other layers on on the ai value chain it's a lot more difficult for european ai companies so sourcing hardware you know on the model side there are there are kind of group of of specialist models um that i think again there is a place for but that those are focused on a specific industry so drug discovery for example um you know there are the likes of base camp research quite high up on the ranking um you know creating muscle models specifically for for drug discovery um just because on for the those companies it's very hard already for them to compete with with the us and china yeah i mean i think the horizontal vertical point is a really interesting one and i'm definitely speaking to investors at the moment who say something similar who say the horizontal play is probably less defensible because of the the sheer size of the budgets that these really big AI labs have to play with and you've got companies like Lagora or Tandem Health who are building very effective vertical plays there's a real moat when you specialize so much in an industry and I think yeah it'll be interesting to see how it plays out for companies for startups in Europe who are targeting more general use cases.

14:56And if you look at the vast majority of the top 10 in the AI100, they're all vertical cases. The vast majority of them are tailored for specific industries. I think that's particularly present on founders' minds and VCs' minds, actually, because a lot of them are scared that one of the big US tech giants is going to come up with a new feature tomorrow. And then that just kind of kills the company that was selling that feature before. And I think that is a lot more prevalent to companies building horizontally than it is for companies really focusing on a specific industry and building in a very granular way for that industry, if that makes sense.

15:35I think you look at enterprise. So that is, you know, to date, the biggest buyer of these solutions and the example of Sana Labs. So, you know, on the kind of horizontal side of things and, you know, for those big tech companies that do have the cash and the ability to kind of build out these big internal systems. With the case of SANA, which is a horizontal play, a kind of one of these sort of emerging AI workspaces. So effectively where you create all your agents, they're all linked up to all your tools within one platform. So that becomes the place you work. That is what they're going for. But in the case of SANA, you know, quickly acquired by a US tech giant, you know, because, you know, they might have been a little bit further behind in their kind of AI development.

16:27I suppose on a kind of M &A side of things, that is a good potentially return for some of these horizontal companies that, again, might find it harder in future to raise money. But, you know, that is maybe the ceiling for some of these companies where, you know, again, they are ultimately building a internal solution and a way of working for these big tech giants, you know, to command that kind of solution and to afford that kind of solution. Those will be the kind of only companies that maybe have the means to pay for that kind of technology, just because, you know, what they're setting out is so far reaching and ambitious.

17:08And I think we talk a lot in Europe about the application layer. And it's something that Europe's definitely very strong on and is building loads of interesting startups in that area. And I think the US gets a lot more focus in terms of companies actually building AI models. But there are loads of companies, especially in the top 10 of the Sifted 100, that are building their own models. Cusp AI, Physics, NeuroRobotics, Black Forest, Labs, Cradle, Poly AI. they're all building their own AI models too. So you have got that foundational research happening in Europe alongside this application layer.

17:46A quick one from me. If your company would like to get a message across to Sifted's audience of startup and scale-up leaders, VC frontrunners and tech advisors, why not consider sponsoring the Sifted podcast? You'll help us interview even more movers and shakers of Europe's venture ecosystem and analyse even more of the most impactful trends confronting them, and you'll reach a hyper-engaged listenership too. For more information, email commercial at sifted.eu. Let's talk a bit about the AI bubble, obviously a trend that we've been following closely at Sifted. The report states that quite eloquently, AI startups aren't going to VCs in this crazy market, it's the other way around.

18:28And a lot of high-profile figures in AI have been talking about the AI bubble for months now, Sam Altman, Vinod Koshla. Many compare it to the dot-com bubble, one of an inevitable crash. What are the symptoms of this that we're seeing in Europe? I think the most obvious sign is the huge increase in capital that European startups have seen coming from VCs this year. Funding has nearly doubled to$8.9 billion this year for AI native companies. and at some point the boom times are going to stop being so boomy. When that happens remains to be seen. I think no one really wants to put their head above the parapet and make any guesses, but some VCs are definitely urging their portfolio companies to go out and fundraise now while times are good.

19:20You've got companies like Synthesia, which is hugely deep pockets, It's raised 180 million in January. According to reports, it's just raised another 200 million. In January, the company told me that it hadn't even touched its previous 90 million raise. So companies are raising while there is money available, for sure. And they're raising as well to beat out competition. I think another symptom of this is obvious, and it's linked to capital, of course, but it's valuations. and Mistral is one of the most representative companies of this symptom I think. Mistral raised back in September 1.7 billion euros at a valuation of nearly 12 billion euros.

20:05It doesn't report its annual recurring revenues, it reports total contract value to date and annual contract value to date which are kind of vaguer terms for how much money you might be making. but their revenues as reported by the FT down more around$100 million a year so there seems to be a strong disconnect between the revenues that these companies are making the cash they're actually bringing in and the valuations they're raising up. And you've got companies as well you've got investors encouraging some portfolio companies to start making contingency plans to start planning for when times are not so good.

20:43One VC advised his portfolio companies to sit down as a leadership group and work through scenarios where the funding market is frozen for two to three years. Because for many of these companies, and almost all of these investors, they've already lived through one market crash in terms of the 2021 post-COVID boom subsiding. And I think it's important to distinguish between the two. The 2021 bubble was a very different bubble to the one right now. The bubble, the AI bubble that we're currently experiencing that is underpinned by tech that is going to change things considerably for years to come it's not just going to evaporate overnight um but some investors still point to moments like deep seek springing up in january as that could have been a moment which could have all of a sudden made people less bullish on the ai market some people say you just need another a big negative headline or some people say you just need a negative headline about nvidia or open ai and that's going to start to shake confidence i agree i think you know a lot of this hinges on on those two companies i think you know in terms of how europe positions itself absolutely while there is money being offered by investors and very freely available you know they absolutely have to cash in in order to stay relevant and and remain competitive with what's going on um across the pond but you know i think on a on a daily basis now what is being reported these big um you know us uh indexes you know a lot is obviously riding on those big companies you know i think i saw something that said 80 percent of the the companies on the smp 500 were actually uh down last week but because of that top 20 percent were were in the green the predominantly predominantly um predominantly tech companies the likes of nvidia for example uh the s &p actually made gains so that can all come crashing down very quickly and that all depends on you know this supposed productivity boom that we are we are going to be getting i think you know a lot of the the use case and applications you know are very exciting i don't think you know we are on the cusp of that burst just yet because i think the promise remains very strong um and people are still very excited by that but yeah as kai says one one headline that is very damaging you know that that is going to have global repercussions on startup ecosystems around the world um you know public markets and and even economies just because there is so much spending you know particularly on the infrastructure side, data centers, you know, that is fueling even more sort of the kind of US GDP growth and consumer spending right now.

23:36And that is a weak foundation to be on because, yeah, we just haven't seen enough tangible improvements yet. We kind of saw this on a lesser extent back in the summer, going back to that figure you mentioned, John, from MIT that was published stating that among the companies surveyed, 95 % reported that their Gen.AI pilots had failed to deliver revenue. And I certainly have not been to a single conference since where this statistic hasn't been brought up. I don't think it feels like it's on the mind of every founder at the moment. I'm curious to know how you've seen startups kind of reconciling this bubble with this number.

24:17What are they saying when you confront them with, you know, what is your product actually doing to help companies make money? And why are Are we seeing numbers like the ones that were published by MIT this summer? I think a lot of founders and VCs would strongly criticize that report. Obviously, it flies in the face of the narrative that there is a market for these AI tools. And there certainly is a market for these AI tools. The big question is, how ready is that market right now? And actually, I was speaking to the founder of a very well-known AI company recently. and they said that they did believe the report was probably accurate.

24:55They thought that it probably wasn't that far off to say that 95 % of Gen AI pilots failed at these big corporates because the majority of these Gen AI pilots weren't being introduced by tech companies, they were being introduced by management consultancy firms. These are, by and large, the big providers of these programs at huge corporates at the moment. And so they thought it probably wasn't all that far wrong. It was just it's the group of companies that are currently implementing these AI solutions. I think we're still in this phase where companies are figuring it out. So obviously, these solutions have been brought to market, which are, you know, changing every week as new updates are brought to the market.

25:41So, you know, that does take time. I don't think it's going to happen overnight. You know, there seems needs to be some level of patience. What needs to remain, obviously, is the deep pockets in order to be paying for this kind of tech. So we, as part of the report, we surveyed all 100 AI companies. We asked them, you know, what are the big blockers to growth? You know, what are the challenges to growing your businesses in Europe? Unsurprisingly, number one was talent. So that was the top. 82 % put it in their top four blockers of eight that we provided them. And it kind of was far that way any of the other blockers.

26:24number two and three which i think kind of relates to what we're saying were gtm challenges and competition and then the third was customer adoption and willingness to pay that kind of suggests you know again these tools have been bringing to market but but customers just don't know how to implement them yet at this stage and if they don't know how to implement them and see wins from them then you know it's going to be very difficult for them to kind of make that roi case and procure that solution. Especially a lot of these application startups have the expectation that they will, you know, these companies will be well-tooled in each of the individual areas.

27:00So, you know, you think about, you know, you've got an AI agent for sales, it's the same for marketing across the different teams in the company. That is the expectation of these application companies. That's maybe ambitious, you know, at this stage for what the kind of level of spend, particularly with the, you know, maybe the scale-up and startup audience. Maybe enterprise looks a little bit different. That jumped out. Obviously, the survey, 70 % of the startup said the enterprise level was their biggest customer group. But beyond that, you know, you're not going to get a huge amount of spend across the company yet, just because, you know, stuff like that put out into the ether does suggest that for the time being, you know, a lot of these processes do need to be human led.

27:40And you've got a customer base in Europe that is harder to sell into. There are situations like Entrepreneurs First culling some programs in Europe and doubling down on cohorts in San Francisco and in the US because it's quicker to sell into customers over there. I was speaking to EF co-founder Alice Benton and she was saying you can get contracts, you can get revenue from US enterprises in like two to three weeks in the US. In Europe, the quickest you'll move is like months and that's fast. Months would be fast in Europe. So I think there's a real schism between the pace of change and the appetite for risk that the group of companies that startups want to sell into has.

28:28So we also asked the question, you know, how many companies would be willing to relocate to the US? We only got that answer as a quarter. I was actually kind of surprised how low that was, to be honest. And I don't blame any of these companies to be thinking about that just because, you know, that's where the money exists. So if, you know, as you say, the appetite for risk is there, the money is there, then, you know, for these companies, they're not as, you know, on a daily basis, they're not necessarily thinking about European tech sovereignty. You know, there are some outliers like Synthesia rejecting Adobe for their acquisition.

29:06But for the most part, you know, these founders are trying to, you know, build their companies, sell their companies ultimately to get an exit and get a return for the amount of effort they're putting into their companies. And, you know, their chances are probably at this stage far higher of doing that by going to the U.S. market. At a conference I was at a couple of months ago, after I mentioned the co-founder of Mistral answered the question about the 95 % that MIT published over the summer. And I think in quite a clever way when he said that the reason the number is so high is because companies are deploying things on a trial, on a test and trial basis.

29:45They're not deploying AI in a structured way in partnership with these companies. And obviously, this is what Mistral is trying to do. they've signed partnerships with some big industrial players in France, including shipping giant CMA, CGM, a$100 million contract over several years to send a team out there to deploy solutions tailored for the company. And Mensch, I think, seemed to be saying in AI, this is the way forward. This is how companies who are offering AI solutions should build their go-to-market. They should work in partnership with these big companies to help them deploy solutions.

30:23Would you say then that it's a problem that in Europe there aren't enough examples of these big industrial companies willing to cash out and sign big checks to work hand in hand with these AI companies to deploy AI? It's not fast enough. It's too costly. What are you seeing in your sectors? Yeah, the market's smaller. The market's much smaller in Europe than it is in the US. And then on top of that, you've also got a market that's more fragmented. It's fragmented culturally. there are language barriers we're starting to see more companies impose english as the language of their office wherever they're based in europe i was speaking to open ai's head of startups for europe africa and the middle east laura moriano a few weeks ago and she was saying that she's seen a big rise in europe-based companies imposing english because it reduces one of these major barriers in terms of language that european startups have selling in their home regions And as well for Europe does have giants, but I think still the first port of call is this mentality.

31:24Okay, you know, what is the US tech solution that I can get into my business? It wouldn't be necessarily to look internally across the continent, you know, something that is much closer to home that maybe I could have more of a relationship with that company and I could build something together with that is fit for my business. I think the sort of the mentality will always be, OK, US first and then, OK, maybe if that doesn't work, then Europe second. And that is a problem for, you know, European industry as a whole. You know, we are not. And same for governments. You know, we are not doing enough to promote procurement from European companies.

32:04That kind of leads me to my next question, which is more of a talent focused one. But I think something we've noticed in the last few months is the rise of a new type of profile in these AI companies, which is the forward deployed engineer. So forward deployed engineer for a bit of context, initially what was initiated as a concept by Palantir. The idea is to have talent that is very engineering focused, but also customer facing that will be deployed within the customer company in order to work with them hand in hand to not only implement use cases that the company wants to see developed, but also iterate and come up with new ideas and new applications for the product.

32:50Here in France, I'm seeing lots of companies hiring for forward deployed engineers. It seems to be the next kind of thing in AI. Would you agree with that? And how critical do you think this trend is? I think it speaks to this growing focus on commercial in AI startups across the region and across the world. The big headline footballers salaries, as reported over the past couple of years, they've all been for people on the research side, people actually kind of developing this like fundamental technology. technology i've been speaking to recruiters over the past several weeks and a number of them have been telling me they expect the next big talent battles to be on the commercial side at these ai companies as the pressure to scale revenue really ramps up from investors and the rise of forward deployed engineers is is part of that it's part of making sure your customer is super happy and feeling really well taken care of.

33:49So they increase spend. I think it's also very telling that, you know, you think about this role, think of it as like an AI consultant, right, that will go into a business and help them figure out how to set up workflows, use AI more efficiently to ultimately improve productivity in order for that company to grow revenue. That is essentially the function. I think what is happening internally in those companies without the role of that forward deployed engineer is right now is, you know, where does AI sit in the company? Who takes the lead in terms of rolling out those changes? It's not like, you know, companies, especially smaller ones, have AI teams at this stage, you know, often, I suppose, taken on by, you know, kind of more data, product, engineering focused people.

34:37But, you know, they might know less about, for example, what sales and marketing need in terms of how they use AI more effectively in their line of the business. So, you know, I think it's, I think it's a good sign of companies being a little bit smarter about how they are employing it, you know, putting their hands up and saying, we don't really know right now. So, so come in and help us. Uh, you know, obviously you need bigger budgets for that, but, um, you know, I think it's, I think it's a promising step. And again, going back to what we were talking earlier about commercialization, you know, whether these AI companies are actually going to be able to turn this hype in, into revenue you know that that kind of has that uh in mind i would say then parallel to that so parallel to companies big big customer companies um still working their way through what ai can do for them 95 figure published by mit we've got we've got european companies like lovable that make headlines because they're reaching crazy amounts of revenue i think so for lovable it was reaching$100 million in revenue over the summer, just eight months after launching.

35:39Should we be skeptical when we read these kind of numbers from companies in AI right now? So, I mean, there's obviously been this big wave of companies being more transparent about what their ARR is. That is ultimately to attract investor attention. We at Sifted run what are called our Sifted leaderboards. So we are collecting revenue data from companies across the continent um some of the biggest companies and most exciting companies across the continent uh a lot of the data that we have to keep confidential um but it's a bit of a treasure trove in terms of particularly on the ai you know you'll get one reported figure maybe maybe in the press um of what a arr figure is but you know when you actually look at the what the actual reported revenue from a filing perspective over say let's 12 months you know those two numbers can be wildly different um obviously ARR is based on the most recent month of um let's say you know for a subscription model you know what uh how many subscriptions have they sold at what cost over the month that is then a projection of the next 12 months to come companies will time that when they've had very good months so you will hear for example when synthecia i've had you know i've seen actually reported on a daily basis so like oh they they sold 1.5 million of ARR in one day and then you kind of multiply that number up by 365 that becomes a extremely big number that can often be wildly different to what the revenue actually collected in from um you know their services being sold over a 12-month period so we have to be really careful and moving on to another trend that is reflected in the report there is more um happening in M &A in AI at the moment in Europe particularly at early stage.