Sandeep Bakshi, Head of European Investments at Prosus Ventures

0:00Sandeep, welcome so much to Scaling Europe show. It is a pleasure to have you. How are you doing today? I'm very well, thank you. How are you? Very good. Very good. And you two know each other, right? Well, I am privileged enough to be working with Paula's partner, Mike. I have not had the pleasure yet to meet Paula, so this is the first time.

0:21Sandeep Bakshi:Yes, I've heard so much. Very exciting to have a live first conversation. amazing very cool well look sandy for those who don't know why don't you give us a quick intro into who you are where you work and your kind of investment piece i guess sure sure so i'm um i'm i look after the venture arm for europe and the u.s for process ventures process is the largest listed tech company in europe so we've got about 170 billion dollar net asset value We are well known for a number of different things. We have a lot of investments in the food delivery space. We've recently announced that we're acquiring in the process of acquiring Just Eat Takeaway.

1:04We have a sizable position in Tencent, a number of different things. I'm going to spare the brain damage of going through all the various things that Process as a group does. But I will focus on Process Ventures. So Process Ventures, we're the venture arm of the Process Group. We're not a fund. We are investors off of our own balance sheet. We sit on about$15 to$20 billion of cash, depending on which transactions are currently in the works and what's going to go through. We're global investors. So I spend my time in Europe and the U.S., but I have colleagues that look after Latin America, India, the Middle East, Southeast Asia.

1:42We're earlier stage investors. So early for us is basically we're investing about any starting at like$10 million type checks. And then we can scale all the way up through our public nature because we're not a fund allows us to invest in companies early and then hold on into an IPO. So, for example, a good example for you guys and your audiences, one of our early investments was into Tencent in China. In the early 2000s, we invested in an all-secondary transaction. We invested$30 million for 50 % of the company. And while that was a successful investment, the real guts was never selling. We sold our first share in the company in 2018 when we sold 2 % for$10 billion.

2:33And we still and that was basically through dilution when we sold that we had about a little north of 30%. And I think today we sit a touch south of 30%. So that's kind of our business. Our model is that we believe that, you know, the businesses that we're backing will compound over time. And we want to be able to participate in that growth as they grow larger and larger. That 10 cent investment must be one of the best returns in venture history. I mean, it's ludicrous. How long are you holding for? Look, I think we have, so one, I have the luxury of not managing that position at all. But we're, you know, that's a position that we're very, very happy with.

3:18And, you know, we're maintaining our stake. But it's something that I don't have any visibility into. amazing that's such a cool um yeah it's an amazing investment to talk about to have on the process type track record a phenomenal uh share and stake and and like you said that you have the flexibility to without having lps you have the flexibility to hold for as long as it makes sense for your business which i think is a really interesting and different approach and model to to traditional fee z funds no for sure that's that's honestly one of the unique things about us is that we try and find companies early, but try and, you know, back them as long as it, you know, makes sense for them as well.

3:57So, I mean, a different example could be Flipkart in India, which we backed, we owned a sizable stake. And when it made sense for them to sell, they sold to Walmart in India. We, you know, we participated in that as well. But it's, so it allows us to be part of companies as they go public or when they sell. And it also allows us to take bets that, you know, theoretically, I let our investment into a company called Oxford Ionics. It's a quantum based company. And, you know, quantum, when I think we were spending time in our investment committee looking at it, they were like, look, this is probably going to be something that we're gonna have to hold sit and hold for a really long period of time.

4:36Lo and behold, it was announced a couple months ago that IonQ came and acquired that company was a pretty healthy return for us. And it made sense for, you know, it's still going through regulatory approvals, but it made sense for the company to be a part of. And again, we're allowed to, we're able to take those really long-term bets, which makes my job pretty fun.

4:56Sandeep Bakshi:How does that change? How do you, how does that change your thesis or how do, how do your investment committees go? Because, um, you know, I, I come from a more classic VC background where, um, we, we also think about long-term compounding, but it's, you know, there is somewhat of, you know, a bias to exit eventually, whereas you could remove that pressure entirely. Yeah, look, I think that what we're looking for, so what it truly means in, you know, looking at folks like, you know, GC, Lightspeed, some of these, you know, multi-platform, Andreessen, et cetera, what we're looking for is really, really large markets.

5:36That's kind of what it boils down to. And now I think that probably the big difference is, And again, you know, it's investing in deeper tech is not a core focus of ours, but we have the ability to do so if we think that it's going to make a if we think that, like, you know, generationally speaking, it could make a difference. So, for example, you know, we announced today that we invest in a company called Cosp.ai, which just raised a fairly sizable RAM. I we haven't announced, but, you know, we're we're we're doing a lot of work in robotics where we made an investment that will be announced shortly.

6:09So I think that taking some capital and making bets that, theoretically speaking, could prove out to be moonshots over time is a way that we allocate our capital. But also, we're also focused on businesses that we think could have a path to exit relatively sooner. But what allows us the ability to do so is that if we see a lot of traction, because I don't think a lot of people are able to predict when there is going to be an inflection point in a business. But if we're able to take some of these longer term bets, who knows what? Like, you know, I don't think that sitting in 2017, people were able to be like, oh, in, you know, at X point, OpenAI is going to release this model and it's going to take off, you know, the way that it does.

6:53it allows us to basically take a little bit earlier perspective on something that could be a little bit longer and then participate in the upside as it comes. And I forget who I'm misquoting here, but it's people overestimate the long-term and underestimate the short-term. I feel like that's a quote that's out there by somebody.

7:14Sandeep Bakshi:I think it's the other way around that you kind of, you overestimate like what you can do in a day or in some shorter timeframe, and then you underestimate how much can change in the long run. Exactly. And so it basically allows us to have, you know, a seat at the table and double down when we feel like the time is right. Amazing. Can we talk about Cusp AI? Because that was like, news came out a few hours ago. You said it's like a sizable round. It's pretty massive, $100 million, I think, Series A or Seed. They're like a Cambridge-based materials company. I think they said that they're building like an AI search engine for the material space.

7:52This is probably one of the UK's biggest rounds of that size at that stage. What about Cusp AI attracted you to it? I think on this one specifically, it's the team. The team and the market size more than anything. We didn't lead this round. We're small participants in it. But for us, again, going back to the point that I made just now, it is such a massive market it's such a strong team and for them to have the capital to really kind of crack this i think is very very attractive for us um and you know we've we've been i think you know one of the we've been selectively making bets on you know infrastructure for either some specific verticals so you know cusp is one corchi is another which we did in the health care space where we think that there are really, really sizable markets.

8:47And it's perhaps a little differentiated or a little unique to build infrastructure for these markets where the larger LLMs or the larger model providers are not yet particularly focused on or may have some either technical difference or regulatory nuance that would make it a little trickier to get into. Yeah, that's a super interesting point. Yeah, because I spoke to the founder of Korsi a few weeks ago. You're right. He's building. It's almost like LLMs, but in the in the medical space. Right. He's you know, they've seen an amazing amount of success. So it's almost like similar with Cusp AI in the material space.

9:23It's like deep LLM tech sort of in quite niche industries, but are still massive, massive markets. Is that like a thesis of yours that you're going to double down on? Are you looking at more companies that are building this type of technology in markets that I guess are underserved by the big tech giants? Look, I think that for us, the way that we think about it is that we're looking at, to your point, right? Niche is perhaps has a misconception of being small, but niches in themselves can be fairly sizable. I think the like first and foremost for us, what we look at what we look for is where can we find, you know, perhaps at the same time or a touch earlier than others.

10:07Teams that we think are exceptional, that's probably the first thing that we look for, first and foremost, in TAMs that are large enough to make our numbers work, to make our math work. And in markets that I think that, you know, well, one of the things specifically when it comes to kind of 40 and vertical applications in markets that are not the easiest to crack. So, you know, so Andres came on and he gave you kind of the story of 40, but they built some very, very specific infrastructure that is very much catered toward their end customer. Right there, you know, dev tool for the health care space that allows, you know, the hundreds of thousands of people that are building applications and AI to actually link into them that they can do so cheaper, faster, more accurate, less hallucinations, more applicability than the average LLM.

11:01Right. You know, you you you read across that for tactile. Right. Which, you know, Paula might be familiar with. But Tactile is going to either they're, you know, they're building. So apart from a lot of the things that they're doing in financial services, they're announcing an agent for underwriting. Right. Which is a very, very specific, nuanced application, which, again, is a niche. It's a massive niche. And everybody that, you know, I've had the pleasure of putting the team at Tactile, Mike, Max and the broader team in front of every customer is like, wow, I can't believe. because a lot of these enterprises don't have the ability or don't have the technical wherewithal or the time really to go out and be like, hey, I'm sure I could cobble together, you know, clay and anthropic and X, Y, and Z tools and build this.

11:50But they're like, oh, there's somebody that's actually done all this hard work for us. And so porty is one tactile. They're more and more that we're finding in these niches where the niches are just absolutely massive. Yeah.

12:01Sandeep Bakshi:Yeah. And maybe maybe like niche is almost more. It's something that's not legible to the broader public. So if you say underwriting agent, the broader public would be like, huh? Okay, like what? But if you think about it, underwriting credit or insurance or, you know, any sort of risk decision is actually happening in most organizations that are somehow, you know, either giving credit to their customers or we're even, you know, trying to prevent fraud as an e-commerce. company, you're always like underwriting a transaction. And I think that's very interesting. Nishes often are not immediately legible, but when you then double click on them, you realize, oh, okay, it's actually everywhere.

12:44It's basically the thing that I love when I'm, you know, spending time in these AI application companies is whenever I have the opportunity to speak with customers or prospective, either prospective customers or existing customers, And this has happened with Tactile, with Cordy, with the other ones that we've invested in, where the buyer shares an anecdote, which is along the lines of my team had been talking about AI or my team had been talking about X, Y and Z. But I didn't think that it was applicable to my business. And then I met. Then I saw this demo or then I saw this use case. And then I was like, whoa.

13:18And then after I, and for existing customers specifically, after I implemented the product or the service, it made my life so much easier. And like that for me is just such a hook because there's a lot of growing pains that companies will face. But that I mean, that's clear evidence of product market fit, which, again, going back to, you know, where we typically look to invest. If the market is massive enough, we'll invest in kind of product market fit where there's just, you know, for I'm, you know, probably going to do myself a disservice, but I'm a door opener. That's where I like to spend time, right, is to basically trying to find the entrepreneurs that are spending so much time building, spending so much time trying to identify their end customer and then try and elevate them to, you know, platforms or podiums where they have the opportunity to sell to the right type of enterprise customer, because that just makes it so much easier and allows them to really accelerate quickly.

14:12Sandeep Bakshi:Yeah, yeah. Okay, elephant question in the room. Do you, how do you implement AI in your own work, in your own organization? This is my current favorite question to the VCs because, you know, we're investing so much in AI. What's your favorite way to use it yourself? So we as an organization have a luxury where given our size and scale, we hired a team that's been working on this since like 2017 or 2018. So we actually have an internal assistant. It's called Tocan that we use. We leverage quite often. It's basically it's in a system that links into all of our internal databases. We also use, I mean, a lot of the call recording stuff that's out there, be it, you know, Granola or others.

15:01We leverage the note taking software that's out there. But I think the thing it's actually interesting. We're an investor in like a in like a private equity business. It's like a roll up type business. and um we were i was chatting with the founder there and the the the interesting thing about venture specifically and you know i'd be curious to hear your perspectives but venture at the end of the day for me is still a deeply human business because at the end of the day it's not you're still trying to convince especially for us where i'm not quite in the needle in a haystack business the the universe of companies that i'm looking for are identifiable like it's not it's not infinite or it's not so much so that's insurmountable for myself and a team to actually go and talk to throughout the course of the day.

15:53For me, the end point becomes, hey, Sandeep, in the universe of, in the infinite universe of capital that exists out there, why you? Why should I give a shit? Like, why are you the right person? And that turns into, it turns into a human conversation. And it turns into like, hey, you know, speak with some of the other founders, speak with some of the other people that are part of our ecosystem. That's where it makes a big difference. But to juxtapose to this other company, one of the things that we're trying to be active on is for ensuring that our companies that we own have enough wherewithal one about how our portfolio companies are deploying AI or telling our portfolio companies how our internal companies are deploying AI and leveraging kind of the knowledge and insight between that.

16:35But the anecdote that I gave when I was starting them on this role of business that we acquired, you juxtapose that with either private equity or some of the different asset classes where it's not a need it's it's it's where there is a larger subset of companies to give you some rough numbers this company that that i'm on the board of we haven't announced it so i'm i can't share what the name is but um last year or not even last year i think in the past six months they assessed 450 000 opportunities to acquire they then send i think 15 000 emails and of that and this is all leveraging some of the tools that are out there with some fine tuning and some business logic and they ended up making four acquisitions.

17:15That is, you know, that's really putting AI to the test. And that's something that's pretty impressive when we see. And we've tried to basically leverage the learnings from that company to actually see what we can do better ourselves. That's insane. That's absolutely insane. And so AI was running that process for them. Is that right? 400 ,000. Yeah, 450 ,000. So it's basically completely running their source of data. That's insane. And then you say, yeah, there's like one in a 100 ,000, I guess, ended up being acquired. Yeah, a little less. Yeah. Wow. That's an amazing example.

17:53Sandeep Bakshi:Yeah, it's like, yeah, there's always like the, maybe not the investing piece that needs to be automated, but a lot of other things around it and things down the line that drive you towards returns. And also, we've been talking a lot in this episode alone, but another episode is about VC value add. And it's super interesting that you're talking about the application of AI in your own internal company as a value that you can give to your portfolio as a leading case of how to use and leverage AI in their own operations, which I think is super interesting. Yeah. Yeah, that's one of the ways that we try and think about it.

18:27really cool um i i also wanted to talk about you know what your what your take is on the state of uh european tech as a whole european venture as a whole you know we're seeing a sort of resurgence in european optimism we had andreas klinger from eu inc on um you've been investing in this space for a while like i guess what do you think european tech european innovation needs to sort of take it to the next level?

18:57It's a big question. Look, I think that from my perspective, having spent some time here, there is no shortage of ambition. There is no shortage of talent. There's no shortage of kind of generational founders that are actually really willing to give back to like other entrepreneurs. It's amazing. And you talk about like, you know, Maddie from Eleven Labs or, you know, Anton from Love of Maddie. I have the pleasure of knowing. I mean, these people are actively thinking about how can we ensure that, you know, this this narrative of, you know, lack of European ambition, lack of European opportunity doesn't exist.

19:40There's definitely no shortage of capital. I know how competitive it is. It's like I wish I wish it was not as competitive. It's an insanely competitive market. I think that the two areas where I try my best or we as an organization are trying our best to be a part of the dialogue is, you know, how can we reduce the regulatory friction if that exists? And, you know, that that's something that needs to be at the highest level, because the reality is, is that for a lot of folks, while there are still a number of European companies, when we want it and the US exists as a massive, massive market, it's we'd love to see, you know, companies have the opportunity to expand across other countries in Europe before maybe maybe making the shift as opposed to getting product market fit in, you know, one one market and then moving to the States.

20:28But at the moment, it's a tough tradeoff for founders to make because they recognize it's like, hey, look, I've got a product here that works and we've got no language barriers, tax barriers. It's difficult to hire, fire, et cetera. There's a number of regulatory friction that exists. And that's an area where we're hoping to be a part of the dialogue and hoping to have we're hoping to have more of a say in that. We're pretty active and we've made some publications on that. I think the other thing that would be great to see more is enterprise adoption or willingness for enterprise adoption for technology.

21:05That's one of the things that like, you know, in the US, there are so many enterprises like I want to know what's going on versus, you know, more often than not, it's sometimes it's like there's some small innovation department in some silo in some part that like, you know, we'll try something. A lot of founders are like, I can't get lost in five years of procurement where there's no case study. Like if like there needs to be a little more from enterprise, at least in Europe, top down where it's like, hey, actually, leaders are very curious about what's going on. And I see that for American institutions, which is probably why for a lot of founders that have just such clear product market fit, it makes a little more sense to go to the States.

21:43It's just like a little more ability to to creep into enterprises would be, especially in Europe, would make it a lot more attractive. Yeah, I've actually heard that from a couple of founders who are selling both European and U.S. companies. It's like the risk appetite within enterprises, the speed that they move, the bureaucracy to kind of get new and innovative tools kind of brought up by the business. But at Sandeep, I'm conscious we are out of time. We're actually over by eight minutes already. This has been an amazing chat. We'd definitely love to have you back on. Thank you. Thank you so much for joining us.

22:16Thank you so much for having me. And thank you.

22:18Sandeep Bakshi:Great to meet you, Sandeep. Great to meet you too. You can jump in person. I very much look forward to it. And congratulations on the platform that you're building. It's amazing. No, thank you very much. All right, let's talk soon. That'll set.