Molten Ventures Plc (GROW) Earnings Call Transcript & Summary

I want to touch on the theme of 20 years. Since 2006, Molten Ventures has seen a lot of technological changes, seen a lot of market changes in that time as well, a lot of geopolitical changes in that dynamic. But one thing to stay true is that we've always looked to support the innovation ecosystem. We've always looked to put capital to work into great founders, great teams and companies that are trying to disrupt significantly large markets and create new markets as well. And with over GBP 1.5 billion invested into that ecosystem, that just demonstrates the commitment that we've made over that period of time. Throughout that time, what has also remained consistent is our focus on portfolio management and disciplined portfolio delivery. So with GBP 770 million of returns coming back through, that just shows you the focus that we put on to making sure we're driving returns for our investors in addition to supporting that ecosystem. We have 85 companies existing within the portfolio now, but there's over 200 investments made throughout that period. So it gives you a sense of the breadth of what we've been able to put together. And we always feel like we're innovative as a firm, and we always feel like having a platform approach to venture capital has been important. And so you'll see that we have pools of capital across the listed plc balance sheet, but also the listed VCT and EIS funds, and we're continuing to expand the third-party capital that we invest. So if we touch on a bit more of that platform, how does it work? The pools of capital have always been important. It allows us to have a deeper breadth of capital to co-invest with. We've always managed this from a governance perspective. So the ability to manage differentiated pools, differentiated investment approaches has been important from a governance perspective, but also to ensure that capital can go into the ecosystem at the right levels, but crucially at the right scale as well. One thing that I think we can accept and what Phoebe will touch on in a bit more detail is that Europe can drive great innovation, can drive great IP, can drive the company creation, but we haven't really solved the capital problem yet. I'm saying yet because there's reasons to be optimistic about that. So we invest across the whole venture capital ecosystem. And we think the ecosystem is deep enough to warrant differentiated ways of investing. So direct investments is the core of what we do. That's why we have a strong team with an investment thesis-led approach. But we also like to invest across secondaries, understanding the time horizons of companies' developments and how they're distinct from GP/LP 10-year structures. Quite often value accrues in the years 10 to 15 onwards. So we'll look to invest in secondaries in direct companies, but also in portfolios of assets that get to that later stage. And that provides liquidity to an ecosystem in addition to the returns that we can drive to our investors. And Fund of Funds has been a crucial part of that. Investing in as an LP into seed funds, in particular, has been critical in terms of driving that ecosystem, but also making sure that we see the opportunities that come through. This is very much still a people-led business, and I think it will continue to be so. Relationships continue to be important in terms of supporting founders being on their boards, but also supporting the ecosystem and making sure that we see the best companies coming through for the next generation. And that portfolio management point I touched on growing the portfolio, supporting the companies, adding additional value beyond just the capital. That's something that we've embedded into our structure and our ecosystem. And eventually realizing returns cash-on-cash. DPI has become the important word for when you're fundraising, the important phrase. But you can see embedded within our system, we've done that for many years. We've always focused on those cash returns, not just at the top winning companies, but making sure as we go through the portfolio, returns are delivered for us, for our investors, but crucially for the founders that have spent many years building those companies. So we thought we'd do a little bit of a time horizon of the 20 years. A lot happens in 20 years, as we know, a lot happens in a week at the moment, particularly in public markets. Stu Chapman, who's going to talk to you later, will give you a bit more of the history of the firm, but founded back in 2006 with Simon Cook. Through that time, we're seeing a lot of technological changes. Obviously, cloud became the incumbent technology over maybe the last 10 years alongside SaaS and enterprise. Finance, London being a great hub for that. Digital assets have also been a really strong theme. But you'll see as we come into talking about portfolio and how we construct the portfolio, that breadth across technology has been an important feature to Molten. But our core approaches remain consistent in how we invest, whether that's direct, whether that's into secondaries or even into looking at the Fund of Funds. We want to understand the long-term structural market dynamics, where does the value accrue? How are the customers buying this product? Why is it going to be 10x better than anything that's in existence already? And technology moves at pace through this period. So we've adapted what we've invested into, and we've adapted to the themes of the market as well. But I think that breadth of the portfolio has been an important factor. We are investing for 8 to 10 years in these companies as an average. And therefore, which of those subsectors are going to be performing well at any one given moment in time is something that you need to balance out across the portfolio. And the slide we have up on the screen highlights a few of the companies that we invested in through that period to give some examples. M-Files that we first invested in, in 2013 is an intelligent information management platform that then subsequently used AI to augment its offering to its clients. And you'll see that thing with quite a lot of the companies. There may be software as a beginning or data at the beginning, that AI being embedded within their product offering is an important part of the defensibility and additional products that they can drive with their existing customer base. We sold M-Files just over the past year. So it was a long journey for us, 10, 11 years of holding that investment. And in a similar way, we invested in Trustpilot. Trustpilot is now known by millions and is a review platform that's scaled over the time that we held that business. And we use secondary to acquire additional stakes in Trustpilot. So by the time it went through its own IPO, we had about a 14% stake and then we sold that down. So it just shows you the ways of creating value along the journeys of these companies is important. And many of these companies are the future as well. They're not just the past. So companies like Ledger, which has been the standard now for digital assets and security and companies like Aircall, which is a communication platform, which is embedded and has grown substantially in the U.S., the CEO putting out just in the last week that they crossed over the $200 million ARR mark. So it just shows you that European companies can grow at scale. This isn't a new theme. This has been occurring for the last 20 years, but I think it's becoming more dominant, more repeatable and you see with the ambition of the teams and the capital that can be recycled that this is becoming a scaled ecosystem. Some of those technology trends that we touched on. We talk about being in the sixth wave of technology now, but this is all opportunity for us. If you're investing in disruptive markets as that pace of disruption increases, that gives us greater opportunity. We are, of course, mindful of a lot of the AI trends, and we'll talk about those in more detail, particularly on the panel later. but it's shifting the way that we think about defensibility and it's shifting the way that we think about how we invest in technologies. But some of those themes we talked about remain true, creating that value, looking at where value accrues, looking at how it's defensible and then looking at the technologies that can disrupt your existing incumbent technologies. So this is an ever-moving picture, but it creates a substantial opportunity. And I think when we talk about Molten and what it offers, it's that insight into those opportunities, that exposure to that growth that's completely distinct to anything else that's in the market. So one of the themes that we've invested in throughout and has been an area where Europe has a real strength, and this is another feature. Where does Europe have a right to win. But in deep tech, and that's covering across semiconductors and some of the companies we've listed here like Paragraf, which is material sciences, and it's looking at graphene for Hall effect sensors, that's a Cambridge-based business. ICEYE we'll hear from later on low earth orbit satellites and SatVu thermal imaging with low earth orbit satellites. These are all deep tech areas, companies like Riverlane, which has an error correction for quantum computing. These are all themes that we invested in maybe 8 years ago, maybe 5 years ago, but these are themes that are now coming to prominence in the markets. And as we're seeing with Europe and the change in the geopolitical systems, that shift to sovereignty of technology, that shift to thinking about dual-use technology, that's become an important feature, and George will touch on that in a bit more detail this afternoon. And as we look to Inga's thematic on health tech, this is an area where huge markets are playing out. We have aging populations. We have budgets that are overstretched. We need to drive productivity into these areas. And so if you look at AI-enabled productivity work streams, looking at things like Deciphex, looking at Aenimal, good examples, these are absolutely crucial. And what we need to have is the adoption within our own health care systems of these tools to create those efficiencies, to create that user experience that we've all become familiar with in our consumer lives. You're going to hear from Clue later on and from IMU as well, great businesses that we've tracked again for some time, and we've put more capital to work in those companies over that period of time. So it's good for you to get updates on where they are and get updates on how they're changing their own ecosystems. And Inga is going to talk you through her investment thesis in this part of the market and looking ahead, how that's going to evolve for the next iterations and again, how AI is inferring some of the thought processes on where we like to put capital to work. So I probably said AI about 54 times already, but you can't get away from it. So the structure of AI that I think we all know and accept is the foundational models at the base layer, if you like, that are driving models of Claude, Gemini, ChatGPT, Llama, everyone is becoming more familiar with these, and they all are improving at a rate of knots. We like to look at the middle layer. The middleware layer is the enabling layer. This is the connective tissue, if you like, between the models and the end-user application layer, the end-user application layer being the utility that drives that user experience. So that middleware becomes the enabling factors. So we start to look at authenticated agents. We're looking at how -- and to decide where compute runs, it's the governance that goes alongside the compute. It's monitoring behavior amongst your teams as an example. So we put here a few of the areas that we're looking at, at the moment that we're excited about. This goes alongside our own thematics of how we like to invest in technology themes. We like to play those enabling layers. We've done that with things like Riverlane, I touched on the error correction for quantum, not picking a hardware layer that we think is going to win, but looking at that enabling layer. In a similar way with Ledger, we touched on this the security layer for cryptocurrency and blockchain. And we'll hear later from SettleMint, which will touch on how they're embedding blockchain in the ecosystem and driving value through their technology. So AI native security and governance, workload intelligence, agentic commerce infrastructure, data infrastructure for AI memory. These are all subsectors of that middle layer that we think are interesting and exciting to explore over the next coming years. I touched on the importance of a diversified portfolio. I think we can all accept that we don't have the magic looking glass that enables us to focus on one area and double down on that. I think the importance here is breadth of opportunity, but understanding that investment thesis going back to that very core investment thesis of value disruption, defensibility and having founding teams that can drive that forward. We've always talked about four broad sectors: consumer, enterprise, hardware deep tech and health tech. But there's the subsectors below that, that we really focus on. And the team has domain expertise in these areas. And then we call ourselves a generalist investor because we'll invest across them, as you'll see today, it's the team that have that expertise in each of their own subsectors that really drives the knowledge, the understanding, the network that allows us to get into quality companies and invest in great founding teams. For the market, thinking around AI and what's disrupting and what's embedded, if you like, I think we've all seen in the last week, a big sell-off in software companies, but this has been happening over the last year as most things that happen with the public markets is somewhat indiscriminate and perhaps slightly ahead of the curve. If you think around software that's domain-specific, if you think around companies that are embedded within enterprises, they run critical infrastructure for those enterprises and they have the underlying data, those things don't get disrupted overnight. And actually, you can, of course, embed new tools, new AI products onto your existing suites. So I think the market isn't yet looking at it with a more nuanced lens than it needs to. But of course, AI is disruptive, and it's a great opportunity for investing in at the same time. There's a debate around how much CapEx that's going in from the hyperscalers into this AI market, how will that get recouped? How will that get recovered? What are the consumer models that will deliver a return on that capital? And I think we can all accept that it's going to play out over a period of time. And therefore, some of the indiscriminate swings that we've seen in the public markets are creating opportunities, of course, but you have to really look a lot deeper into the underlying models of those businesses to understand what is their defensibility, what is their own opportunity. So the nice thing about standing up here every year is that we get to talk about what we did in the last year, and we're very busy at Molten, you'll see that across the day. So a lot of activity. We've continued to have new direct investments. We've continued with our secondary investment strategy, and we continue to drive returns in the portfolio. Some of those, as you'll see with Revolut and with ICEYE are partial exits as we start to portfolio manage our positions. We always think around the plc balance sheet, GBP 1.4 billion of value and things like Revolut and ICEYE and the rest of the companies in the core represent around GBP 900 million of that. So balancing that portfolio, being sensible at the time to take some liquidity, but riding sufficient value for the upside is something that we've always focused on, and we think that's the right way to approach any portfolio and deliver value and proof points for our investors. So Phoebe is going to go into more detail on the European opportunity. You can see the market has been scaling roughly sort of 20% a year since 2016 when we first IPO-ed. We can see that the deal count numbers do move around. Some of the high peaks of value have gone into those later-stage rounds. But what we're certainly seeing is that Europe, despite having scaled in the last 10 years, is still about 3x more than the U.S. market for capital. And so with the same number of companies and the same IP, we clearly need to drive that capital coming through. But what we're also seeing with the shifts in geopolitics and the shifts in the underlying scaling that we need to see from our companies, the need for sovereignty, the need to focus on dual use, that's driving not only GP and LP capital coming into the market, but it's driving government capital coming into the market. And that institutional capital that's coming from pension funds is being driven by governments as their own agenda as well. So it's becoming a part or rather a greater part of industrial strategy. I think I would argue that post Second World War, it's always been a part of industrial strategy and go into some of the history of 3i perhaps, but those are the rationales. And so this is an opportunity for us. Where we're seeing success in the market. We're seeing the capital get recycled. We're seeing the entrepreneurial talent start new businesses. We're seeing the investor ecosystem expand and the knowledge within that ecosystem expand. So we're starting our own flywheel effect. And you could argue the U.S. had that experience in the 1980s when they reformed their own pension schemes to invest in this market. So I think the opportunity is very real, and Phoebe will give us much more detail on that in a moment. So this is the -- what we said we would do slide and focusing the business on the core areas where we've always invested, A and B investing, particularly on Series B, where there is a gap to capital. There is a need for deeper pools of capital, but that's where our expertise lies. Scaling the portfolio, not just from a number of company perspective, but growing the assets and having third-party capital that can come alongside is absolutely critical. And then we're talking to the balance sheet strength. That's what we're here for. That's what we do. We want to make sure the balance sheet is strong. And as we recycle capital, we have that NAV-accretive approach to investing and recycling that capital. So I'm pleased to say we've demonstrated a fair amount of success in those areas. The amount invested is to the half year, GBP 50 million. That's probably going to be nearer to GBP 90 million, GBP 95 million by the year-end. We continue with third-party strategies around a Series B focused fund, dedicated capital to that part of the market and also Molten East, which is a new Eastern European strategy that's coming close to getting a first close. We hope both of those will get a lot more traction within this coming year, and we're able to announce to the market that progress. And then realizations, as I say, over GBP 100 million just in this financial year coming through. And we've been thinking about NAV accretive use of that capital. Every incremental pound, what's the best use of that versus investing, be that direct or in secondaries, what's the best use of that versus buying back our own shares. And so we've committed GBP 50 million already executed to date into share buybacks. That's NAV accretive. It reduces our issued share capital and drives value to our shareholders while the public markets aren't reflecting the full value of the underlying assets. And we have an extra GBP 10 million going into that program now. So just to finish, looking ahead, we're very ambitious as a group. We have a very strong team here. We've been adding to that team, and I hope some of you will get a chance to meet Franco, who I think he sat upstairs. And he's joining us with over 20 years of experience in this market, and we have progressed our team as well, internal promotions with George and Inga both moving to partners. You'll hear more from them today, and you'll see very clearly why that was a good decision. And expanding our pools of capital is absolutely crucial. But we have a platform play. We have the governance structures. We have the institutional base to take on more capital. I always say we are not opportunity constrained, just capital constrained. And therefore, the opportunity to do more with more is absolutely there. And we continue to support that European ecosystem. Fund of Funds relationships remain important to us. You'll see some of our fund partners on the stage later today, and crucially delivering strong returns for our investors. So we're really delighted to have you all here. I hope you enjoy the day. Do stay around. And hopefully, we all learn something as well. Thank you.

Hi, everyone. My name is Phoebe. I'm a Principal here at Molten. And today, as you can see, I'm talking about the opportunity in European venture capital. As we reflect on our last decade as a listed company, I wanted to take a look at the broader European ecosystem, where we've been, where we're headed and most importantly, what's next. So a big theme of today is reflecting on Molten's journey since our IPO a decade ago. And as we look back 10 years, European venture looked markedly different, as you can see in this slide. So at the time, there are 47 unicorns, the largest being Spotify at $8.5 billion. There are now over 400 unicorns, the largest being Revolut, one of our portfolio companies at $75 billion. This has resulted in a globally leading ecosystem, where London, for instance, is now the world's fourth largest venture hub outside of -- only beaten by 4 U.S. cities. This transformation has created a critical and scaled asset class. So annual capital invested has now increased over 3x over the last decade to over $70 billion, where the tech sector now represents 15% of European GDP. As a result, venture in Europe has evolved from emerging to really institutional grade with returns continuing to be on par with what we've seen in the U.S. So why has this been the case? So this didn't just happen by accident. Europe has several enduring structural advantages that have only compounded over time. The first is talent. So this is both structural and intentional. So Europe in and of itself is an attractive environment to build a company. We are home to a large and diverse home market with 500 million people holding great global political and economic influence. We have also long been a market where migration is actively encouraged on a government level, such as the skilled worker visa in the U.K. or the EU Blue Card. And investment in technology talent is no different, where our workforce has been uniquely positioned to succeed. We are highly technical. So there are more PhDs working in European deep tech companies than in the U.S. We also have 40% more software developers than the U.S. And the success of the broader European venture ecosystem has also bred success, where VC-backed companies have become more attractive for skilled operators to come to Europe and build global businesses. So to R&D. So a lot of the R&D strength is -- well, talent is only as good as the R&D that often supports it. And as you can see from the chart here, Europe is home to more than half of the world's top science hubs, and it has continued to remain the home of innovation for the last few decades. This has promoted deep academic and industry collaboration. Unlike in the U.S., where talent typically stays local and pulled into big tech, across Europe, this is fragmented and deeply tied to our academic roots. This is also expected to continue in AI, where we're home to some of the world's most renowned AI research hubs such as Oxford University and ETH Zurich. And as Ben said, this has really established a sizable flywheel over the past decade. So to walk you through this slide. So one is the success stories. So Europe has definitely proven we can build global winners. As I said, we have over 400 unicorns with nearly 3x that amount achieving $100 million plus in ARR. Fortunately, a lot of our portfolio companies are also leading this charge, such as Revolut, Wise, N26, Ledger and Aircall. And these aren't just European success stories, they're really global ones. And this has brought an exceptional talent density, which is distributed across numerous hubs. So what's interesting about Europe is 90% of European unicorns have actually remained headquartered within Europe. And this level of innovation, particularly from a distributed perspective, is unmatched globally where numerous European cities that have the same amount and depth of talent is typically seen across 2 to 3 U.S. hubs. As a result, this has accelerated both capital and expertise throughout the ecosystem, where we're seeing the development of second and third time founders, people who learned at Revolut, Spotify, Wise are all spinning out and founding their own companies. For instance, Revolut alone has had 50 companies spun out of them. And this has recycled capital to make real returns. So over the past decade, over $900 billion has been realized in exits across European venture, reflecting our ability to move towards an entire ecosystem built on the maturation of venture versus just the success of 1 or 2 individual companies. However, as Ben was talking about, Europe is really at an inflection point. So having built an ecosystem over the past decade, Europe now sits at a once-in-a-generation convergence moment, creating an unprecedented opportunity for European tech. This is largely down to the sixth wave of innovation. And maybe to walk you through some of the innovation cycles that came before us. So the first was the steam train in the 1770s, then railways, heavy engineering, mass production and finally, IT in the 1970s, all of which were the cycles and timeline between each of those waves had compressed significantly. And now this new wave is no different. So the era of AI, automation and robotics, largely marked by the launch of the first public GPT in 2022. This has disrupted and created entirely new categories and business models, which at the time were not previously possible or conceived and where there remain no clear winners. This has been combined with technological and also structural changes, where for the first time, European start-ups can compete globally from day 1. So on the technological side, GenAI as a technology, when combined with falling cloud costs means that generational start-ups can truly be built from anywhere, which benefits Europe given our fragmented ecosystem. And also structurally, which are some of the benefits that we've seen both in the U.K. and the EU, where the EU in particular supported harmonized frameworks such as the 28th regime, incorporating EU Inc., the AI Act and the Digital Markets Act. Yet the need for self-resilience remains ever important in an increasingly complex global environment, where European companies can be the beneficiary. Here, there remains a real opportunity, in particular, to rebuild critical tech infrastructure, particularly across real assets such as AI infrastructure, semiconductors and batteries, creating the need for European champions in addition to global ones. So on to AI, obviously, a big topic for today. So unsurprisingly, given the disruptive nature of this technology, funding in the space has accelerated drastically where we have already seen the emergence of category-leading European champions, both at the foundational and application layers. These rounds not only attract the most capital, but also premium valuations, typically 50% above what we're seeing for non-AI companies at the growth stage. And at Molten, we're excited by a variety of use cases where we focus on driving differentiated insights through our thematic-led approach. So just to name a few high-level areas that we're thinking about. So one is the need for new infrastructure. This is both true on the cloud perspective, both for SaaS models and also the emerging needs of AI compute, particularly around inference chips and the ability to utilize and the high utilization of energy workflows. The second is augmenting existing workflows. So we're seeing this in next-generation vertical software when combined with open finance has a greater ability to have personalization than what we've seen before. The second is the rise of agentic commerce and consumer, and you're also seeing this in personalized medicine in health care. And finally, is unlocking new categories of spend that were previously too difficult to access. So resilience in dual-use defense tech is an obvious example of this, but we're also seeing this in other industries such as within financial services, the focus on illiquid assets and wealth management. So looking ahead, we expect enterprise adoption and budgets around AI to continue to grow with success ultimately compounding around the winners who are able to demonstrate real productivity and ROI over time. So to deep tech. So given Europe's historic strength in academia and our skilled technical talent, our appetite to invest in frontier technologies isn't new, but the capital required to succeed often is. So for instance, Apple alone spends more on R&D than the top 10 European tech companies combined. And at Molten, this is an area we have long been excited about, backing leaders in the space such as ICEYE and Riverlane across quantum, semis and space. And now our focus is on the exploration of further frontier technologies, particularly in areas where tech sovereignty is required to scale, such as dual-use defense tech. And here, we are exploring three unprecedented waves, which are creating a golden 2- to 3-year opportunity before the market normalizes. The first is Europe is spending more on defense than any time since the Cold War. We're also having governments moving at massive emergency speed with sizable leverage effects. And the last is private markets flowing countercyclically with defense stocks outperforming with sizable capital now flowing into the space. However, whilst there remains opportunity and ambition to succeed, the access required -- the access to capital required to do so remains a key challenge and one as an ecosystem that still needs to be addressed. So obstacle one of what we're facing today in Europe is the scale-up gap. So as you can see, Europe has approximately the same number of start-ups raising for the first time from VCs as the U.S. However, at the breakout stage, this reduces by 2x, further compounding at the scale-up phase where there are 7x fewer $100 million rounds per year than the U.S. This misalignment has also created an opportunity for U.S. investors to increasingly play in Europe, leading 60% of Series C investments versus 20% at the seed stage, at precisely the stage where capital is needed the most to scale real global businesses. Here in lies both our obligation and also opportunity as a European ecosystem, where more funding is required to build both European and global champions of tomorrow. And the second, as Ben discussed as well, is the sort of why is this gap so prevalent in Europe historically. The lack of European long-term institutional capital has historically been a key barrier for European venture, where more involvement is required, particularly given the need for increased sovereignty as we think about the current geopolitical environment. And on that note, when we think about sovereignty, the latter isn't about -- it's not about protectionism, but it's about winning through innovation and performance as part of creating and the formation of a new world order, which are really capitalizing on our structural advantages that we already have today to build both European and global businesses. And on the positive, there really have been improvements in particularly public funding into venture. So we are seeing renewed commitments from an EU perspective through the European Competitiveness Fund and also on our home soil via the BBB's industrial strategy. However, there's also a need to unlock numerous other stakeholders to provide long-term capital required to succeed. And part of this is cultural and also institutional. So for instance, from a public procurement perspective, only 9% goes to innovation in Europe, which is half the level of the U.S. The same is true on a corporate level in terms of how our corporates actually interact with start-ups. So only 20% of European corporates actually interact with start-ups versus 50% in the U.S. And the last is patient capital. So pension funds and university endowments make up 40% of European venture funding versus 90% in the U.S. And this gap in particular, has to change, particularly as we think about the need for capital, both at the growth stage and R&D-intensive industries. In the U.K., we've had Mansion House reforms, which have gone some way to changing this, but we further need to capitalize on this momentum given the importance of the next few years as an inflection point. So how can we, as Molten win and capitalize on this? So with the support of our investors, we believe we have built a differentiated offering to help address these challenges and also back the new champions that Europe has offered. So we play at this from the early stage through access via Fund of Funds, which is confined on-the-ground local ecosystem winners where we can get access to their breakout stories. The second is from our direct capital at both the early and the growth stage, both areas where we've demonstrated a track record over the past 20 years, Series A, B and C, supported by our thematic-led sourcing and investment thesis approach and also in the U.K. by EIS and VCT capital. This continues at the late stage where we can be flexible such as buying out prior funds or accessing secondary opportunities, where the depth of our network in the U.K. and across Europe has also enabled us to support companies into the pre-IPO and beyond. And I think this is a great slide just to reiterate really the depth of our network within Europe from both the portfolio and Fund of Funds perspective. And this has really compounded and driven our own flywheel of connectivity, relationships and expertise, enabling a differentiated sourcing opportunity across the continent. So to wrap up, just say last few words. European Venture has come a long way over the past decade, but the opportunity ahead remains even greater. As the next wave of innovation unfolds across AI and deep tech, Europe's strength in talent and research and also prior success stories enables us to position us globally. At Molten, our job is to back these generational leaders, leveraging our full cycle platform, deep network and thesis-led approach. And you'll hear today more from our portfolio companies and the rest of the investment team, how we're doing that. So thank you.

My name is Dr. Inga Deakin, and I lead on health tech here at Molten. I'm going to talk about the sector, some exciting trends and how our portfolio is benefiting. At Molten, we define health tech as companies whose customers sit within their health care ecosystem. That includes health care systems, consumers, pharma, medical devices and life science tools companies. Historically, these have been quite distinct markets with different buyers, different sales cycles and regulations. But the boundaries are changing. Consumer wellness companies are pursuing clinical impact. Medical device companies selling continuous glucose monitors are adding patient engagement software. Pharma companies, particularly with the rise of GLP-1s like Ozempic are adding patient engagement software, and they're adding coaching and direct-to-consumer models. Big tech is all in. Google, Amazon, Microsoft, NVIDIA, all building significant health care capabilities. And in the Frontier AI labs, OpenAI and Anthropic have both launched dedicated life science and health care teams, and they're hiring aggressively. This convergence creates an enormous opportunity for companies positioned at these intersections to grow and define new categories. There are many ways to evaluate health tech companies. Ultimately, the companies are improving access to care, efficiency of delivery or product development, clinical outcomes and/or the experience of everybody involved. As in any sector, we're looking for an outstanding team, a large market and best-in-class product. And in health tech, we're also interested in two things here in workflow integration, companies that embed into clinical or pharma workflow, deliver immediate ROI and become really sticky and a clear data story. So companies with unique often ongoing access to data that compounds their advantage over time. One important note, we invest in companies that help pharma, but not the drug itself. So infrastructure, platforms, picks and shovels, but not taking the therapeutic risk. I'm going to talk about three megatrends in health tech and about our portfolio. First, of course, is AI, which is being rapidly adopted. This is data from the American Medical Association from last year, showing a steep growth in adoption to now 66% of physicians. I'm sure that's out of date, they're much higher now. And what's striking is how fast value is being created. OpenEvidence is effectively the ChatGPT for doctors trained on medical scientific literature. They've grown really fast, over $100 million revenue, free to use for doctors and about 40% of U.S. doctors are using it every day. And you can see a $12 billion valuation. Note the differentiated data source. Companies like Abridge did not exist a few years ago. Now they're valued over $5 billion. Their revenue went from $7 million in 2023 to over $100 million in mid-2025. That's 15 to 20x in 2 years, less than 2 years. And their valuation doubled in 4 months last year. This is AI to capture clinical conversations and it fits seamlessly into workflow. Anima is one of our own fast-scaling examples. They built a clinical operating system for primary care, which automates triage, documentation and workflow. You know what it's like contacting a GP. In 2025 alone, they processed over 5.7 million patient records -- patient requests, forgive me. That's over 15,000 a day. The median response time is 1.2 hours and over 90% of requests are resolved the same day. They're now deployed in over 600 practices, meaning about 10% of the U.K. population has access to one or more of their products. Contracted ARR has grown to over GBP 20 million quickly because this ROI is so fast and significant. This is what workflow integration looks like at scale. And we led the $12 million Series A a couple of years ago. The second area is in pharma and precision medicine. You may remember a fantastic exit from Endomag breast cancer company for over $300 million a couple of years ago. These companies are speeding up and improving drug development and ensuring the right treatment gets to the right person at the right time. There are many high-growth companies in precision medicine, and a great example is Tempus AI. They IPO-ed in 2024 with about $700 million of revenue. Now they're guiding to $1.3 billion this last year, 2025. That's 82% growth. They combine precision medicine testing with AI-driven data products for pharma, which is a great example of those changing categories. Genomics testing companies such as BillionToOne show that new entrants can grow quickly and Natera and Guardant are showing this growth at scale. So genomics, perhaps the first wave of precision medicine measuring DNA and RNA, and we believe immunology can be next and potentially even more significant given how central the immune system is to so many diseases. Here are three of our portfolio companies, three angles on pharma and precision medicine and none of them are drug developers. IMU Biosciences has an AI platform mapping the immune system at unprecedented resolution. They have fantastic access to data and are becoming a discovery engine for precision medicine products. We led the Series A, and you'll hear more from John shortly. Deciphex is Ireland-based, built by a team that really knows what it's like to be a pathologist. They provide digital pathology software, AI tools and services to both pharma and clinical customers, including the NHS. By delivering the service, they generate a unique and ongoing flow of pathology data to better train their models. That's the data flywheel in action. They're now at over EUR 20 million revenue, and we led the Series C to support their expansion. PolyModels is in the infrastructure layer. ModelFlow is the digital backbone for pharma process development, helping scientists design, simulate and optimize manufacturing processes. Clients report a huge reduction in the number of experiments they need to do, and we led their Series A last year. All three of these are picks and shovels. None of them, as I said, are taking therapeutic risk, but they are rapidly advancing the rapidly growing field of precision medicine. The third trend is for -- is in tools for patients to engage with the health and for people. The wearables market is large. OURA, who make the biometric rings, raised $900 million at an $11 billion valuation a few months ago. And the CEO said to set a new standard for what wearables can achieve in advancing preventative health, which is a really clear example of those boundaries blowing. Notably, the European digital health coaching company, Oviva, recently passed EUR 100 million in revenue and raised EUR 200 million to scale. Consumer-first offerings like Neko in Europe and Function in the U.S. have raised big rounds and are scaling fast. Also public companies like Hims & Hers. So consumers and patients are paying health care systems are recognizing the benefits of investing in preventative and chronic care. There's a long way to go, but we're backing some of the pioneers. We call this thesis consumer and preventative health because we don't think of consumers as just healthy people tracking their wellness, many have or are at risk of chronic conditions. So our portfolio spans this spectrum. Clue invented Femtech. They literally coined the term. So today, they have over 10 million users, and we're going to hear more shortly. Hilo delivers clinically validated blood pressure monitoring at the wrist. 30% of adults have high blood pressure, 50% are not aware. And of those 50% that are aware, 50% are not managing their blood pressure well. Hilo bridges the medical-grade precision with consumer accessibility. We led their Series A, and they raised a $42 million Series B last year. PocDoc is focused on cardiovascular risk. 90% of these diseases can be prevented with early detection. This product -- their product includes a finger-prick blood test, which can give you -- read your full lipid panel and give you a 10-year cardiovascular disease risk score in under 10 minutes. So this is preventative care made accessible. There's so much more we can talk about. I'm really excited about the flow of talent into health tech, bringing tech and company-building experience from other sectors into health. And at Molten with our interdisciplinary approach, we're well positioned to know what good looks like. Our portfolio is already improving millions of lives and millions more to come in the future, which is personally extremely motivating and translates to growth. So I'd like to thank you all for your support. And next, we're going to hear from two of our portfolio companies that I've mentioned.

I'm going to give us quite a quick overview of how we see the world within energy and energy tech, and I'm going to do that predominantly through the medium of pictures and graphs. I'm going to try and sort of distill some of the investment opportunities and areas that we've been focused on at Molten, specifically around certain areas of software and AI and data and analytics businesses. But the idea ultimately is to tee up a couple of the presentations that you're going to hear directly after me and give you a bit of a flavor as to why we invested in them. So the first thing to say is I very purposefully called this section energy tech. Had I been doing this presentation 2 years ago, it might have been called Climate Tech. Had I been doing it 10 to 15 years ago, it might have been called Clean Tech. And had I done it in the period prior to then, when I started my career in this space, we probably just would have called it renewable energy. And in some respects, that gives you important points of context. But in some respects, as I'll go into, it's largely irrelevant. So if I have one slide, which tries to capture where we are in the debate right now, it's this one. So as mentioned, in the recent past, this probably would have been about net zero pathways and carbon emissions. But right -- in 2026, the discourse at the moment is about energy independence and energy security. And ultimately, where you don't want to be is where Europe is right now, and that's reliant on fossil fuel imports from countries that you probably no longer want to be reliant upon and they've got a long way to go there. And the best way to build out energy security and energy independence is to build and own your own generation, i.e., renewables. And the reason I put this slide in is a sort of counter to some of the discourse that you might have heard around people rolling back from net zero is that periods of energy security concerns have always driven the fastest decarbonization. So that was the case in the '70s, the '80s. I think it was Carter that popularized the term. This is around quarrels with the Middle East and OPEC, and it was that which drove much faster transition than any concerns around carbon or anything else. So this is important. This is helpful. But for me, this is far more important. And these two charts that I've got here are, a, probably slightly unnecessarily complicated looking; and b, frustrated my compliance team immensely in trying to verify them. But the picture that we're trying to show is quite a simple one, and that is every country is somewhere on this journey from bioenergy to generating clean electrons via the trough of burning hydrocarbons, so burning fossil fuels. That is the megatrend, that is undebatable, that's inexorable. And it's been this decade to the 2020s that have been the pivotal period accelerating that transition. And again, that acceleration doesn't really have a lot to do with carbon. It doesn't have a lot to do with energy security. It's just economics. So we entered this decade and the cost curves for the dominant technologies within renewables, so solar PV and lithium-ion, they got crushed such that renewables just became the cheapest form of energy generation. So again, anything you might hear about people rolling back from net zero and carbon not being a thing, 2024 and 2025, somewhere between 93% and 95% of all new generation that came online globally was renewable. That's just a function of economics. As with any market that inflects and these have been massively inflecting markets, consensus, i.e. market analysts are really bad at forecasting markets that move so quickly. It's a job that I used to do, so I'm allowed to say that. But this has been a period of exponential growth, and it's continuing to grow a lot faster than people expect. It always exceeds consensus forecast, which get upgraded every year and never quite go far enough. And that's not slowing down. That's not changing. The CapEx requirements for the transition are in the trillions. And if anything, the risk to those numbers from where we sit now is only to the upside. And that slope might not look particularly steep. But what that's showing you is the incremental demand, power demand just from data centers because of AI from 2025 to 2030, that's equivalent to like the whole of generation of Japan, the whole consumption of Japan. These are huge numbers. The CapEx related trend is a megatrend. So what does this mean for us as investors? What is evident, there's clearly going to be lots of hunting grounds at the earlier stage in the growth stage just because of the scale of investment, the scale of market change and the scale of disruption that this all causes. But as Ben alluded to earlier, as early and growth stage investors in the equity layer, you really have to think about where value accrues. And over the last 20 years, I've spent a lot of that time looking at this space, you probably wouldn't have wanted to be investing in the hardware, just commoditizes, probably wouldn't want to be investing within new materials. The dominant materials just become more dominant as they traverse down their cost curves. But what has emerged and is emerging are really interesting data and analytics businesses and software businesses that enable a lot of these capital flows, but also benefit from the new market structures that it creates, the fundamental transformation of energy and what's going on at the moment. What do I mean by changing market structure? At the very high level, energy markets are just becoming more complicated. Data requirements are increasing, and they're becoming more localized, more real time. And I've sort of proxied that theme by looking at German power prices, which is a little bit niche. I'm loving the laser pen as well. This is basically showing you as renewables grow within the mix, greater intermittency, greater price volatility. This is just one way of showing that renewables and the new system creates a lot of extra energy data, a lot more complexity. What does it mean at a broader level? The thesis -- and again, it's difficult to do in 10 minutes, but at the high level, the move from a fossil fuel economy to a renewables economy is equivalent of going from analog to digital. So analog fossil fuel world is very few centralized points of generation, unilateral grids, relatively stable data and relatively stable and predictable data requirements. And the use cases can and have been served by relatively rudimentary software businesses or in the most part, consulting and services businesses. That's fundamentally different to the renewable world where you now have millions upon millions of distributed energy resources, incredibly high data frequency. So these markets are real time and incredibly dynamic to try to show the price volatility, all needs monitoring, all needs optimizing. The data types of how the markets evolved fundamentally changed, lots of multimodal markets, and it's two-way. The grids are now two-way as a lot of people know. So the use cases in order to scale up renewables because this is ultimately a capital issue. These markets have to financialize. So everything that you might need to do, price forecast, risk manage, ensure, these require much greater levels of advanced intelligence. This is an area that we've been looking at for a couple of years. And there are, as I mentioned, increasingly valuable software businesses that are emerging in this space because the ones that have emerged from this space aren't fit for purpose over here. So even in the last 24 months, we've seen GBP 1 billion-plus exits in the U.K., in Europe and globally for businesses that are building and starting to grow in that space, and it's been a focus for us. Again, quite high level, but what defines in this world, some of the businesses that we like. We love software businesses that do what good software does, so embeds in workflows, eats workflows, gets adopted by the industry. Can you -- as a data provider into energy, can you get integrated into contracts? Are you bankable? Can people underwrite against you? Can people lend against you? If you are -- the output of your business is a price like it is with Neil's business that you're here or a forecast or a benchmark, this is the case with Quentin's business, does it become a trusted reference point in the industry like all good information services business become? Because if you do, become incredibly sticky and can build a very valuable moat in that business. And then as I said in the middle are the clear acquirers in this space because of the change and because of how disruptive this new energy system has become, there are lots of acquirers, very acquisitive space around energy data at the moment. So a few portfolio examples that tie into some of the work that we've been doing here. Those of you that were here last year might remember hearing from Sightline Climate, which is building a data and analytics platform for this new economy. From BeZero Carbon, who I think were also here last year. They're building a ratings business but focused on carbon markets and new ecosystem assets, are equivalent to a Platts or a Fitch. And then in this section, you're going to hear from General Index, which is building the world's first technology-based price reporting business focused on commodity and new energy markets. And then from Quentin, who has already built one of the global leaders in terms of benchmarking, forecasting and valuing these new renewable energy assets. So as I said, I've actually got 30 seconds to spare, so I've sort of done my job and not overrunning. It's very quick. So anyone can find me afterwards or feel free to e-mail me, tell me that I'm wrong or if you want to chat about any of this stuff, we can do that. Otherwise, I'm going to hand over to one of our portfolio companies next.

The title of this panel is AI Hype, Substance and Opportunity, which we hope really encapsulates an awful lot of what we're thinking about as we navigate this kind of this really compelling wave. It's clear that AI is kind of both everything and nothing. We've talked a lot about it already this morning. No question, it presents extraordinary opportunity, but we're also aware that it's also creating extraordinary levels of noise. We've seen unprecedented levels of capital coming into the space. We can all see companies that are scaling at speeds faster than any of us have seen before. And on the other hand, there's a huge amount of volatility. A single product announcement, DeepSeek a year ago, Anthropic last week can hit the public markets and swing software stocks 30% to 40%. So today, it's not about is AI important? I think that debate is sort of fairly settled. It's more about just discussing with experienced investors, how do we separate the signal from the noise? How do we think about where durable value is being built and created here? Where are investors focusing or indeed getting carried away. And so to kind of dive into that, I'm thrilled to be joined by three very experienced investors, as Ben has said, all of whom we have had the privilege of backing as Molten as Fund of Funds partners of Molten in Pietro, Max and Audrey. So I might ask each of them to just give a brief intro to yourselves and your funds, and then we will kick off.

Yes. I start.

Please.

Thank you. Thank you. Thank you, Nicola. It's been great to be here. I'm Pietro Bezza, I'm Co-Founder and General Partners at Connect Ventures. Connect Ventures is an early-stage venture capital firm. We are seed specialists. We are venture generalists in terms of the areas we cover. I personally have been investing in B2B enterprise software and AI for more than 12 years. So this is a very special and exciting and challenging times. Before being an investor, I was on the other side of the table as a founder and CEO. I co-founded a company in Milan in 2001, speaking of platform shift and bubble sign, there was a dot-com. And yes, we got acquired like 2007, like 6 months before the iPhone came out, and we were fundamentally a mobile business. So navigating the platform shifts and the hypes has been something I've been done for more than 20 years in tech at this point.

Max?

I'm Max, Co-Founder and GP at IQ Capital, which is a deep tech VC firm based here in London and investing across Europe. We're actually turning 20 years old this year as well. We've been focusing on deep tech for...

We share the birthday.

We're about 6 months younger, so as it should be. So yes, we've been focused on deep tech from the very beginning. We have backed about 100 companies over this time. And AI is a theme across our core focuses and we -- we actually have a few co-investments with Molten within that. And so we'll be very keen to pick up on some of the trends that we're seeing and talk more about it.

Yes. I'm Audrey. I'm a partner at Tapestry VC. We are a U.S.-based fund that invests globally, and we specialize in pre-seed and seed, and solely focus on founder archetype. So we'll exclusively invest in repeat founders. So the way I like to think about it is any pre-seed stage fund needs focus to filter the world for the best founders. Some folks do that based on geo, some folks do that based on sector. We really focus on founder archetype and look for founders who have had experience building great companies and we'll back them on their next journey. So we'll often spend months, if not years, with these founders thinking about what their next journey is. The fund started 8 years ago. It's a really small team. It's 4 of us. And before that, I had spent basically a decade in Silicon Valley for studying and then starting my career more traditionally in investment banking, left that, came to London, started a company in fintech out here. When that shut down, I reconnected with my former colleagues who are starting Tapestry and have been basically building the fund ever since.

Thank you. Max, I might start with you, right? We'll take it sort of from the macro to start with. I mean I mentioned at the beginning, we are seeing unprecedented levels of capital coming into the space, both from venture capital and also the large tech giants investing hugely. And all of this at a time of heightened sort of macro uncertainty. We're seeing those tensions show up in the public markets with these kind of big volatility swings. Question to you, I mean, are we in a bubble? And when we look at the level of investment, how do you see how all of this will sort of ultimately drive returns?

It's a good question. I will try to be focused in kind of giving some perspective on this. So I guess one perspective is that disruptive tech always create bubbles. The Gartner Hype Cycles have been well established. The question really is whether or not this bubbles burst or whether they deflate or sort of grow into themselves because the ROI case empowers or underpins the valuations. So on the valuation side of things, there are two perspectives or two angles. One is the other sectors that AI is actually impacting. The SaaSpocalypse, there's even a word for that now that we have seen last week, has at one point, wiped out some 300 billion of the valuations. It has now recovered, but it shows that there's a lot of nervousness as to how this technology will affect various sectors, and it doesn't stop in software. Arguably, if you are an agritech, you should be thinking or could be thinking whether and at what point AI will be affecting your models. So it's actually a broader story for many industries. So for AI itself, fundamentally, it's all about ROI and ROI in turn, does depend on whether or not there's the fit for purpose of the technology itself. Given the costs of AI, which we see now a short-term perspective for significant reduction, both on CapEx and OpEx, the only ROI that can pay back for it is complete replacement of human labor in certain functional areas. That is ultimately what will drive the tools cost from $10,000 a year to engineer to $150,000 per year instead of the engineer. And we haven't yet seen that replacement. So the question is whether or not the market will actually have the patience to wait until that happens and whether that happens quickly enough for the market to not lose their patience. That said, at the moment, the business use case is still being searched. There are some good areas like coding and legal tech. There are lots of areas like AI agents where it kind of works 90% of the time and then it doesn't, and that makes it unusable for the business applications. So there's still a lot of search. And I think the jury is out there whether that will happen in time. The model development race still continues. We will see the news every week when they beat each other and it still is one of the highest depreciation asset on earth of all time. So the cost is there. On CapEx, I think the interesting bit is, of course, it's a huge investment, but it is powered by equity, by all the hyperscalers. And that is a big, big differentiation to what we have seen in many other technologies in the past in that there's no leverage to unravel the valuations. Ultimately, if Microsoft or Google takes 5 years longer to return their investment than initially anticipated, what's is their call. Maybe their stock will come down a bit, but ultimately, that's an interesting aspect. The OpEx side of things, I think there was this race for developing the best model. Now people are increasingly understanding that it's actually inference costs that you can't walk away with. Even if your model is greatly fit for purpose for a particular application without continually investing into inference, you are going to be second to the least less good models.

Inference costs, just maybe to -- explain to the audience, when we talk about the training of models, which has been kind of all of the time and the data that has gone in over the last few years to get these models to the level that they are now when Max is referring to the inference cost, it really is kind of practical -- the training is about recognizing the patterns, the inference is, this practical application of these patterns of these models into you data to sort of infer the results that we're all looking to see.

Yes. And put simply, it can scale quite substantially in terms of costs. So the more -- if you think about your ChatGPT or any other model that you're using, it's how many of those experiences that you've had specifically, dialogues does it actually remember. That all scales up. People talk about it as kind of the new marketing budget really for many of the applications. And they are very significant. So that's all back to the ROI question. So the final point that I'll make is just one or two kind of historical parallels in that we've kind of both seen it and not seen it before. Yes, there are some of the learnings from the dot-com era that are informing where we are perhaps. And while AI in all likelihood will create a bigger outcome that we can possibly imagine now, in the long run, whether or not it will probably not get -- it will probably take longer than the 2 or 3 years that everybody is talking about and what the market probably has the patience for. That said, the multiples for most AI firms are actually not outrageous or not as outrageous as we saw in the dot-com bubble. Many of them are 25, 30, 35x multiples on revenues and those revenues are growing quite rapidly, but they are not that durable is the test of time. And finally, one, if some of the performance are correct and we see, say, 20% of the world GDP being transitioned on to AI, we're talking $25 trillion. If it's a 2x that some others are also talking about, that's $250 trillion. And that's why the race is there. There's a lot of money at stake, and we have not ever seen a technology, which is not just a tool used by humans, it's also able to function on its own independently. And increasingly, we're going to start seeing that, which creates effects which are very difficult to predict.

Yes. And all of us individually, collectively and as a set of investors are now navigating our way through -- I mean, our jobs are to place bets in this space. The space is that's so deeply complex with such a degree of uncertainty. I mean, Pietro and Audrey, interesting, Max, I hadn't heard SaaSpocalypse before. But we spent the last 15 years through cloud transition, assessing a sort of SaaS benchmarks for what does best-in-class actually look like. And these next generation of AI companies are ripping that up, this kind of triple, triple; double, double, double as your framework to 0 to $100 million ARR. We're seeing companies today going from 0 to $10 million in year 1. But we're also seeing -- coming back to the durability point, these AI companies with high churn with a lot of testing. There's a lot of people trying to figure it out. So much higher churn than traditional SaaS. You both -- maybe Pietro to you first, you're both -- you're investing at sort of the earliest stages of these companies. How are you thinking about the companies that will sustain, the companies that have true defensibility and durability? And what's the kind of framework you're using to assess that?

Yes. So as pre-seed and seed investor, we -- it's a bit different situation, right because we do the initial investment and initial assessment, most of the time pre-product and all most the time pre-revenues. So we don't have the chance to assess the quality of the revenues. There are no revenues at all. But what we always have is the founders, right? And so what we need to do is distinct those high-quality founders that are a good fit for the AI for those founders and start are more like riding the hype and they have no right to win in this new world. And what we're actually seeing, I think it's cool to share is that as we meet more and more founders, we call next-gen founders like really native, people that spend only 5, 10 years in their life as operator, as a founder or as in start-up, only fundamentally working on AI, they think radically differently. And so the realization is that lot of founders are not living in the future or living in the past or even in the today. And what means is that they are not optimizing their product decision and the go-to-market decision on what will be the future, what will be the tomorrow way of working. And so for us, the main question that when we try to do the assessment on the founders like are they building for tomorrow? Or are they building still for today? Are they led by the fear of AI or they like fully enthusiastic and really embracing AI. And so they fundamentally make product decision on where the models are going rather than building for the constraint of today. That is a radical difference in the way you think about the product and how also you design your organization, like they are hiring agents like literally like they're not hiring people, 2 or 3 engineers and then they think they can really, really serve thousands of customers. It's quite a radical change. So what I'm realizing is as a founder today in AI, you can't be half-pregnant in AI, like you have to be all in, take your risk and take your opinionated bets on how you design things. And so for us, now the exercise is like, what does that mean? How can we identify and select those founders that can actually -- they are building for tomorrow. Is it quite a new thing? Then, of course, we also are investing in our portfolio company at advanced stage. Then in that case, we have revenues and we have assessment on the durability, the sustainability of their market share, et cetera. This single framework that we have is the question that we ask ourselves is do this company can grow when their customers -- when its customers, became way more efficient, but also way less numerous. And so that is going to happen. We might not know when, but I think it's going to happen. Company -- AI provides much more output per person. And so eventually, company can do -- will do more with less. So it was much more, less people. And so if your business model and what you're serving as a software company today, an AI company today is people fundamentally is you might be more in trouble and your revenues might not be have good retention.

Audrey sort of on that theme as well. I mean, has the way you diligence companies or founders changed as you sort of transitioned into these investments?

I think it's a good question. Some of my answer will echo what Pietro has been saying, which is we invest truly pre-revenue. So the thing I'm assessing is founder ambition and execution capability. And that has not changed. The bar is high as always. Speed has changed. So how quickly can you get product to market, how quickly can you test product market fit, iterate, pivot, that's at an unprecedented level. And that's what I'm doing in terms of diligence, but then you invest in companies and obviously, they need to go raise the next round, right? And so we spend a lot of time with our portfolio companies thinking about what will the next stage of diligence look like. And there, I think -- I mean, I was just in San Francisco maybe 3 weeks ago, and I was speaking to a Series A investor who literally told me, we will not take a meeting with the company if they haven't gone from $1 million to $10 million in revenue in a year, right? And I mean that's rhetoric, and I think that, that rhetoric is great. There was a Bessemer report last year that said there's a difference between shooting star companies and supernova companies. And every founder read this report and started freaking out about how I need to get to $100 million revenue in a year, right?

And there are good storytellers with you.

Phenomenal storytellers. I think the issue with some of that rhetoric is it, to your point, doesn't talk about revenue quality. That's great for prosumer, but namely a GovTech business that's sold $10 million in a year. And so I think where it does change is how we speak to our founders about what the next round looks like and what are the metrics they will be assessed on. When we're investing it so early that some of this hyperscale is not really in our realm of diligence, yet it does really affect how we think about the next round.

And on that point, I mean, you both talked about focusing on sort of founders and the profile of the individual or the founder who will be able to navigate this depth of uncertainty that we have ahead. I mean, how do you -- because there's so much uncertainty, do you take a more sort of -- is there particular sectors that you have built thesis on that you believe are more ripe for disruption? Do you take a more targeted approach to either sectors or indeed kind of layers of the stack? I mean where are you placing your bets today? I mean, Pietro, maybe...

Yes, we need a compass, right, for. And so again, our main thesis about founders. But of course, we need to be put our mind on specific sector or corner. So one actually we spent more time recently is everything about robotic and physical AI. So we like try to go as vertical as possible, as narrow as possible and making founders with extreme domain expertise and a strong right to win in specific angle. And one that we think there's a lot to build and a lot to innovate is the full infrastructure layer for physical AI and robotics, which is way less hyped than the digital AI stack. And there's much more to build, and there's a massive gap in tooling and infrastructure. So one of the recent investment we made at Connect is called Mosaico. And Mosaico, it solves a big problem, which is try to fix the data infrastructure for robotics and physical AI, which is currently quite broken. So when you want to train an autonomous system, a robot, you need to collect a lot of sensor data from cameras and LiDAR. And these data are very heavy, very diverse. And in order to generate the training for the autonomy, you need a massive management of data. And normally, as of today, the models companies, they also need to build the infrastructure of data to serve this purpose, which doesn't make any sense, like? And so what Mosaico does actually has built a purpose-built data management platform, purpose built for the physical AI. And we think that is a massive opportunity because that -- once this is fixed, it will unlock much more innovation and much more acceleration in the generation of autonomy. This is a good example of how we are to deep dive in the vertical.

Max, you guys are also kind of quite focused on the deep tech space and infrastructure there. Where are you focusing your time? What are you excited about?

Well, to the extent that we're staying within the domain of kind of AI side of things. So one area for us is crossover into the big verticals like life sciences. You've heard from IMU and John just before this, and that's our latest investment actually. So super delighted to partner there. And we have done a few others in sort of drug discovery into cancer therapies. We see the opportunity in bringing robotics into that domain. And then echoing also on the physical AI, it seems that robotics and industrial automation is just starting to scratch the surface of what's possible. And you can get sort of on one hand, one of the biggest challenges for us, given the sort of size of our funds and the stage where we invest, we can't pick a fight with the top 5 model providers, and they keep increasing and improving the functionality. So how do we invest in spaces which are big enough to build a big outcome, but defensible enough for us to see that technology mold and to do that. So [ Magentic ] here in London, for example, is focusing on helping robots train much faster. So you show it once and it can repeat that task so that you don't have to do the reprogramming. Vsim sort of similar to the kind of foundational models in robotics, it's okay, you've collected all the data, how do you then train the model itself, it's kind of spinoff from NVIDIA that does it 100,000x faster than NVIDIA's own product and so on. So that domain of industrial automation for specific AI applications or very deep areas which require a lot of specialist knowledge like life sciences are the two things that excite us a lot at the moment.

Audrey, what are you most excited about?

Some of the similar themes, but one that comes to mind immediately is where do you -- where can you have physical moats. And so hardware is something that I think has been historically out of fashion inventor over the last decade. And I think that's -- there's a big resurgence, whether that's in robotics with physical AI or even in consumer, one of our portfolio companies, which some of you might actually own product from is a company called Nothing. They make best earbuds, right? They do literally earbuds and phones, right? And if you think about the real estate of your pockets, you have two back pockets in your trousers. And so AI will need to be conveyed into our experiences through a physical product. There are many companies iterating on what that looks like, whether it's visual, auditory, wearable and Nothing is ideating in those spaces as well. But as of now, it's earphones and physical phones, and you normally have 1, maybe 2. And so to me, that's a really defensible kind of mechanism.

What's really interesting is that there's sort of a consistency of physical AI, data infrastructure, hardware, we haven't really talked much about sort of the application layer. I mean I was going to ask the question like where will you not invest? I mean coming back to maybe this question about the hype end of our title here, what are -- what are you actively avoiding? Where do you not want to invest? You touched on defensibility, perhaps areas that the LLMs that kind of we're seeing such reach here. Maybe speak a bit to where would you not place a bet today?

I can -- my like short and fast answer is probably marketing tech just because you see so much changing in that space. You've got -- we had SEO, we're moving to GEO. How do we think about ads within model providers? That's a space that I think is just too early from the stage that we're investing in right now to see something defensible. Similarly, I always will ask a founder, what does your product experience look like when we have Opus 7 and GPT-9, right? And if there's no clear thesis of how this product experience gets better as models become more performant, which then lends itself to AI wrapper style is...

Defensibility point, which is obviously top of mind for everybody. Red flags areas you won't invest?

Well, I mean, as a deep tech investor, we wouldn't invest in an AI wrapper company. We wouldn't invest in a SaaS company at this time. We look for that defensible moat and ability to really capture a big chunk of the global market through that. And I wouldn't put -- I wouldn't take a bet against the top 5 foundational model developers either. So it's kind of carving out, as I was saying, those areas where you require specialist expertise to do that and you are not likely to be affected by these models as they go. One thing I will say is that even on the foundational model side of things, historically, in two funds ago, you will actually hear from that company later today, Causalens, you have kind of an angle at a very, very specialist model because all the reach in the market has been about LLMs, yet LLMs are a very small element of what is actually possible in AI. And they're not a very good solution for many of the problems that we're actually trying to solve in the industry or specific applications. So we're expecting to see a lot more intelligence from those models and more intelligent decision-making, digital scientists sort of -- and some of the AI labs are trying to do that. And actually, these are the sorts of things that we're seeing in Europe as well and maybe it's intriguing whether or not it's an investable proposition, but some of these things are super cool.

So clearly, the AI is changing the economics, right? So there are a few categories or a few areas that are suffering and they are not exciting. But -- and probably the horizontal software, the workforce software, those are clearly suffering mostly in the markets in the recent market downturn, but also objectively and intellectually like they start losing sense in an AI first world. But on the positive spin, the trigger or new effects that are brought by AI. As an example, coding now is free, like AI is commoditizing the -- what was used to be not like an important capability was creating and building software. This is free. And so as a consequence of free code, now there is an incredible abundance of code. And that triggers new needs and new opportunities. So more code, it means more surface area of code to be secure. So security AI powered is still a very interesting area and it's going to here to stay. Again, more AI, more data, more data pipeline. So the data infrastructure, I think has a structural tailwind also in the future. Developer tools, we can discuss, more code requires more code to be handled to be secured to be debugged to be deployed. Agents are already doing that. But still, it will require still toolings for handling such a vast volume of code. Paradoxically, now I think company will compete more on what to build rather than how to build. It's more important to decide the direction and the strategy of what you're going to build rather than just the ability of shipping code. And so at this point, design and user experience become a differentiation point more than before because it's when you decide and you explore what to build. And so I think tooling there are augmenting the ability for exploration, for creativity, for designing, for building great user experience, I think that will have a much bigger adoption.

Okay. We're nearly on time. So let me sort of finish up with a quick question. A very brief sort of one answer -- one word answer. We -- there's very clear optimism on the sort of trajectory and the direction of travel and the opportunity for AI. Are we in a bubble today that will correct? Yes or no?

Yes or no? Yes.

Yes, by most measures, we are in the bubble. Will it burst? I don't know.

Everlasting bubble?

I think financially, yes, but technologically not.

Financially, yes, technologically not. Thank you, all three of you for taking your time to come and join us today. It's been a pleasure to have you and really appreciate your time and your insights.