The European Startup Scene

Summary

Podcast Summary: "The European Startup Scene"

Software Engineering Daily | May 26, 2026
Guests: Edward Keelan (Octopus Ventures)
Host: Elena Baroda

---

Episode Overview

This episode delves into the maturing European startup ecosystem, focusing on technical founding talent, the role of AI in reshaping software, and what it takes for Europe to compete globally. Edward Keelan – a partner at Octopus Ventures with deep experience in enterprise software and AI – shares candid insights about founder qualities, capital and ambition, technical debt, AI disruption, regulation, and engineering culture across Europe.

---

Key Discussion Points & Insights

1. Introduction to Octopus Ventures

• Background: Octopus Ventures manages ~$1.5B, with a focus on B2B software and enterprise AI, active across Europe and Australasia, but with a strong UK angle. ([02:45])
• Investment Scope: Writes checks from £2M to £20M, from late seed to Series C.
• Mission: Invest in ideas, industries, and people that can "change the world for the better." ([02:45])

---

2. Qualities of Great Founders

• Three Main Criteria:

  • Founder–Problem Fit: "Why this founder for this problem?" ([05:23])
  • Ability to Motivate: Clarity in communication is vital—for both customers and teams.

    "If they can't explain it to me, chances are they're not going to be able to explain it to their customers or future employees." – Edward Keelan [06:15]

  • Trust and Openness: Prefers founders who call him first when things go wrong rather than sugarcoating issues.

    "The worst thing is where you've got a founder who basically is trying to sugarcoat everything to you. Because as an investor, you can't then try and solve those issues." – Edward Keelan [06:42]

• Technical Founders:

  • Avoid "boxing" founders (e.g., only technical vs. commercial).
  • Technical obsession with product can be a pitfall if practical application or real problem-solving is missing ([07:15]).
  • Currently an "engineering uptick wave"—engineers are driving new opportunities, especially with AI:

    "If you're not backing founders who have a strong leaning toward product and technical development, then you're going to lose quickly." – Edward Keelan [08:56]

---

3. European Founders and Talent Ecosystem

• Universities: Not a lack of talent—universities like Oxford, UCL, and Cambridge produce technical founders, but commercialisation lagged until recently ([09:29]).

  "The companies coming out are much more commercially minded with this technology than historically." [11:11]

• The Flywheel Effect: Success begets success. Big exits and visible role models (e.g., Spotify in Sweden, Revolut in the UK) inspire new founders ([11:41]).
• Geography: Talent is widely distributed—across Eastern Europe, Spain, Portugal, and the UK ([12:13]).
• Location: London is a springboard—English language, access to major markets, and proximity to the U.S. ([13:06]).

---

4. Capital and Scale Constraints

• No Big Capital Shortage: The real constraint isn’t funding but scaling ambition and building enduring companies without quick exits.

  "The biggest problem within Europe is our ability to take early stage startups, keep them in Europe and then build them out bigger and turn them into the next Spotify." – Edward Keelan [14:43]

• Exit Culture: European founders tend to sell earlier than U.S. founders; changing cultural and risk appetite ([16:01]).

  "European founders will tend to exit much quicker than a U.S. founder." – Edward Keelan [16:16]

---

5. The AI Boom: Opportunity and Threat

• AI versus Previous Hype Cycles: Unlike crypto and metaverse, AI companies generate revenue quickly ([19:10]).

  "The AI boom is completely different... These companies are generating revenue really quick." – Edward Keelan [19:25]

• Sense of Urgency:

  "If we haven't started doing it already, then we're too late." – Edward Keelan [20:57]

• AI-Native as a Criterion for Leadership: Actively replacing "old school" CTOs with AI-native technical leaders ([21:50]).

  "If that is not where your brain is... you're going to very quickly lose, you're going to be overtaken." [21:59]

• AI Disrupting Incumbents:

  • Softmageddon: Public SaaS companies vulnerable to AI-driven alternatives ([22:29]).
  • Building defensible software: Key lens—Can an LLM easily replicate your software? ([23:37])
    • Verticalized, application-specific AI is more defensible than horizontal tools.
    • Creation tools, discovery tools without proprietary data, and generic workflow tools are increasingly at risk ([24:50]).

---

6. Funding Criteria in the AI Era

• Polarized Funding Environment:

  • AI-native companies: Growth at all costs, metrics like gross margin often ignored ([27:45]).
  • "Rule of 40," burn multiples, gross margins still matter for vertical SaaS or slower-growth businesses ([27:45]).

  "If you're adding £1M AAR, the most you can really afford to burn is £1M at this stage... Even that's high." [29:30]

• Efficiency is Expected:

  • "It's no good to just hire endless engineers anymore... You should be much more efficient."

  "If you aren't, then that's a question for management: what are all these people doing?" – Edward Keelan [30:20]

• Impact on Graduate Engineers: Reduced hiring and opportunities as AI replaces entry-level jobs ([30:54])

  "Graduates are struggling to get jobs. The number of jobs for graduates has gone down by 25% over the last 12 months." [31:17]

---

7. Can Europe "Win" in AI?

• Defining Winning:

  "Does American society win because Elon Musk is a trillionaire? ... I don't think that's great for society." [32:51]

• Europe's Opportunity: Pooling talent, fostering cross-border collaboration is key. European universities improving; Sweden cited as a role model ([34:08]).
• Large Language Models: Should Europe build its own LLM as a defensive move? Unclear ([35:08]).

---

8. Policy and Regulatory Landscape

• EU Inc. Initiative:

  • 48-hour, 1-euro company incorporation across the EU; modeled after UK’s friendly environment ([35:08]).

  "You can start a company in the UK in an hour with a bank account... some [countries] you have to put tens of thousands of pounds into bank accounts." [36:04]

• What More is Needed:

  • Calls for more founder-friendly institutions, risk-reducing support (e.g., French programs covering founders' costs) ([38:15]).
  • It's not a lack of founders, but a need for better support institutions and ecosystems ([38:15]).

• Does Regulation Slow Startups?

  • He hasn’t seen it as a major constraint ([40:27]).
  • Internal technical issues like monolithic architectures and technical debt are much greater challenges ([41:14]).

  "If I looked at what's been the biggest engineering issues that I've had in my companies versus regulation, the internal issues... are way greater." [41:14]

---

9. Technical Due Diligence & Debt

• Learned Importance: Technical DD now a key focus—looking for scalable codebases and clean technical history ([42:20]).
• Technical Debt Perception: Now a strong red flag; in new (AI-native) companies, less of a problem if just built ([43:56]).
• Founder-Led Go-To-Market:
  • Dangers: Chasing large but tangential deals = product drift and loss of product-market fit ([45:14]).
  • Discipline in focus essential to avoid Frankenstein products.

  "I'm very nervous about companies that are trying to boil the ocean and serve many, many sectors and geographies early on." [45:22]

  • Engineering is now central to GTM (Go-To-Market), with product/engineering driving possibilities in AI ([46:47]).

---

10. US vs. European Opportunities for Engineers

• Stark Contrast: US startups offer higher risk/higher reward, but also instability. Europe’s scene is maturing, with more stability and the promise of spawning future giants ([47:48]).

  "It's unbelievably brutal... European startup is probably much more secure in terms of job security." [48:09] "I would encourage [engineers] to stay in Europe for that reason." [48:36]

---

Notable Quotes

• "If they can't explain it to me, chances are they're not going to be able to explain it to their customers or future employees." – Edward Keelan [06:15]
• "The companies coming out are much more commercially minded with this technology than historically." – Edward Keelan [11:11]
• "If you're not backing founders who have a strong leaning toward product and technical development, then you're going to lose quickly." – Edward Keelan [08:56]
• "Our ability in Europe to hold on to the ownership of the companies, to build them much bigger, so they build those beacons for other entrepreneurs... that's what's going to inspire the next layer." – Edward Keelan [11:41]
• "The biggest problem within Europe is our ability to take early stage startups, keep them in Europe and then build them out bigger and turn them into the next Spotify." – Edward Keelan [14:43]
• "If we haven't started doing it already, then we're too late." – Edward Keelan [20:57]
• "European founders will tend to exit much quicker than a U.S. founder." – Edward Keelan [16:16]
• "It's no good to just hire endless engineers anymore. Why are you hiring all these engineers? What is wrong with your code?" – Edward Keelan [30:20]
• "Graduates are struggling to get jobs. The number of jobs for graduates has gone down by 25% over the last 12 months." – Edward Keelan [31:17]
• "Anything we can do here to make it so much easier to incorporate and unify companies would make a huge difference across Europe." – Edward Keelan [37:08]
• "If you're not, then that's a question for management: what are all these people doing?" – Edward Keelan [30:20]

---

Timestamps for Important Segments

• [02:45] — Edward introduces Octopus Ventures & fund focus
• [05:23] — What makes a great founder
• [07:15] — The strengths and weaknesses of technical founders
• [09:29] — Universities and commercialisation in the UK/Europe
• [11:41] — Flywheel effect and regional role models
• [13:06] — Is London the default? Distributed talent
• [14:43] — Capital versus ambition and scale constraints
• [16:01] — Cultural reasons for early exits
• [19:10] — AI's real revenue impact versus prior hype cycles
• [21:50] — Preference for AI-native leaders in portfolio companies
• [23:37] — Will AI eat all software? Investment thesis shifting
• [27:45] — Growth at all cost is out, unless you’re AI-native
• [30:54] — Graduate opportunities, hiring pressures
• [32:51] — Can/should Europe "win" and what does that mean?
• [35:08] — EU Inc. and why easy company creation matters
• [38:15] — What else Europe needs for founders
• [40:27] — Regulation vs. genuine engineering constraints
• [42:20] — Technical due diligence in practice
• [45:14] — The founder's effect on product-market fit and engineering
• [47:48] — The European vs. U.S. startup path for engineers

---

Episode Tone & Takeaways

Edward Keelan brings candid, practical optimism coupled with a VC's realism and caution. The conversation is open about Europe's challenges—ambition, cultural risk, technical legacy—but positions the continent as poised for a new wave in the post-AI software world. The message to founders and engineers: Europe has the talent and the capital, and with sustained ambition and supportive policy, it can build truly world-class companies.

---

For founders, engineers, or investors interested in Europe’s evolving tech scene, this episode offers an unvarnished, insightful, and hopeful look at where Europe stands and where it’s going in tech innovation.

Transcript

[Narrator]

Europe's startup ecosystem is maturing rapidly with companies like Revolut, Lovable and Ligora demonstrating that world class technology businesses can be built and scaled on the continent. While the US remains the dominant force in venture backed software, as home to the largest markets, the deepest capital pools and the most ambitious exit culture, a growing number of European founders are choosing to build at home. Edward Keelan is a partner at Octopus Ventures, one of Europe's largest and most active venture capital firms where he has spent over 16 years leading the B2B Software and Enterprise AI fund. His portfolio spans seed through Series C with a focus on European founders building in AI, vertical SaaS and enterprise software. This long view experience gives him a rare perspective on what it takes to build enduring technology companies in Europe. In this episode, Edward joins Elena Baroda to discuss what separates great founders from the rest. How AI is reshaping the software landscape and threatening established players, the state of the European startup ecosystem and what it needs to compete globally and what engineers and founders should be thinking about as the industry enters a new era. Alena Baroda focuses on GTM for developer tools and AI startups with experience in observability and building tools for MCP servers. She is based in Berlin.

[Elena Baroda]

Welcome to Software Engineering Daily. Today's guest is Edward Keelan, a partner at Octopus Ventures. Great to have you on the pod, Edward.

[Edward Keelan]

Yeah, thank you so much for having me. It's a real privilege to have been asked to be on the podcast, but also really nerve wracking as well. I feel like, I don't know, a few people sort of frowning at VCs already and engineers wondering what it is that we actually do. So yeah, I'm slightly nervous because we're going into different worlds than I'm used to.

[Elena Baroda]

I think you'll be great. We already had a short chat before, so for those who don't really know Octopus Ventures, can you tell, give people a bit of background. You're a very active, one of the most active VC funds in Europe. You are also one of the largest VC funds in Europe and you deploy around like 270 million USD. So for you it's in pounds, 200 million pounds a year because you're based in London. So tell us a bit about that. You've been there since 2008 and you've been leading the B2B software fund. Tell us about yourself. What got you into VC and what is Octopus about?

[Edward Keelan]

Yeah, I won't cover all 18 years. Otherwise that will probably take the entirety of the podcast and nobody Needs to hear that. But yeah. So Octopus Ventures, we manage around about one and a half billion dollars. We're part of the Octopus Group. The Octopus Group is a B corp and sort of our purpose is really to invest in the ideas, industries and people that are going to change the world for the better. So that was kind of the overarching ethos. We do that in all sorts of ways. Lots of people know us because part of the Octopus Group is Octopus Energy, which is the UK's largest energy supplier. Interestingly, we actually started as an investment company, not an energy company. So I've actually been working at Octopus for a lot longer than the energy company was set up, but it's been a huge. And so many people know the brand for the energy company. But Octopus Ventures is completely separate to that. And as I say, we invest everywhere from sort of first check fund all the way through to Series C, and we do that across Europe and also Australasia as well. So we've got investments all over the world. My personal focus. So I look after the team that runs our software fund. So we have a. Used to be called a B2B software fund. We're sort of going through a kind of branding crisis. So we probably call ourselves B2B Software and Enterprise AI Fund, because most of what we do now is Enterprise AI. It's a 600 million pound fund, lots of dry powder, super active. We're trying to do a deal every single month, if we can, and trying to find the best opportunities that are out there. So any engineers or founders out there that are looking to raise, Love to talk to you. We do everything from kind of late seed in my team through to Series C, so checks from sort of 2 to 20 million pounds.

[Elena Baroda]

Fantastic. And where do you look for founders? Like, can anyone apply from anywhere in the world or are you geographically quite constrained?

[Edward Keelan]

No, absolutely, anyone can apply from anywhere in the world. I mean, for my fund in particular, really is partly because of the expertise we hold. We generally has a UK angle, so there has to be generally some sort of UK angle. Doesn't necessarily mean to say the head office, but where the companies are expanding into the UK and they've got UK employees and revenue, I think that's really where we're best positioned to help companies as well. So we do have companies in lots of other jurisdictions, but we do have this kind of prominence to the uk. But anybody can buy. Please do. Probably get, I don't know, 10 pitch checks a day. Love talking to founders, love finding out what's out there. And Sometimes I think VCs can seem like a bit of a black box. But the most important part of our job is finding the best founders. That is literally the point we can make. Our most critical decision is the point in which we invest. And so origination of founders is basically critical to being a good vc.

[Elena Baroda]

And how do you define what is the best founder to you? You said it's all about finding the best founder.

[Edward Keelan]

Yeah. The human side of me really struggles with this because I'm not somebody that likes to judge other people, especially on early judgments. I always find that really, it's a really weird one. In my outside of work life, I think I'm a very, I like to hope I'm a very non judgmental person. But then obviously in my working life that's kind of part of the gig. So for me there's kind of a number of things. I think number one is like, why this founder for this problem? You know, why is this founder the right person to solve this problem? I think that's a really interesting question to ask. Number two, maybe the most important of all of them do I think they'll motivate both employees and customers. And a lot of the time that motivation comes from their ability to easily explain things. If I think back to my worst investments I've made, this is probably where a founder has come across so bright and so intelligent and throws tons of stuff at me and I'm like, I don't understand what this company does, but that guy or girl clearly knows more than I do and therefore I'm going to back them only to find that the trouble is no customers understand what they do and the employees don't necessarily fully understand, understand the direction they're going. And so for me, communication, they don't have to be like the kind of archetype founder that's sort of all extroversion. You can have very introverted founders, but with a very clear purpose and explanation of what they're trying to do and the problem that they're trying to solve. Because if they can't explain it to me, chances are they're not going to be able to explain it to their customers and future employees. And then just my final point would be somebody that I can trust. I like to think that for my founders, if things go wrong, I'm the first call they make, not the last. That's the relationship that I want with them. Because the worst thing is where you've got a founder who basically is trying to sugarcoat everything to you. Because as an investor, you can't then try and solve those issues.

[Elena Baroda]

And are there any specificities, anything that is even different when it comes to technical founders where you think this makes them the best?

[Edward Keelan]

I think as always, we like to put things in boxes, if that makes sense. We like to say this is a sales type founder, this is a finance type founder, this is a technical type founder. And some of the best commercial founders that I've come across are actually very, very technical because they deeply understand the engineering problems potentially they're trying to solve and what they're trying to do for engineering teams potentially. So I'm always careful not to kind of box off them. But it is true, and it's a bit cliche that with an engineering finder they're obsessed about product and not thinking about the commercial opportunity. That can be really problematic. And we see that quite a lot, especially with kind of university spin out founders where maybe there is a research product they're trying to do something crazy with. I don't know. We got it a lot, actually. We got it a lot during the metaverse kind of phase. I feel like there's loads of people that were doing lots of research around how the metaverse can be used for this and that. And you like the practical application of that. And actually whether that's solving a real problem, I think is generally where technical founders struggle. They start from the product and not the actual problem. I do think that however it goes in waves and right now we are on an engineering uptick wave. It's the engineers who are dictating to the people, if you like, the wider people, what the problems are that are now going to be solved by AI and how those application layer products can really kind of change the way in which we work. I think it's almost a very weird, it's a very different world to say the world of vertical SaaS which we were in just five years ago, where actually engineering was not necessarily the game changing thing, because everyone knew what a vertical SaaS solution looked like, knew how to develop one, et cetera. We're in a very different world now where the premium on not just engineers to build, but engineers to help the rest of us understand means that if you're not backing founders that have a strong leaning toward product and the kind of more technical side of the development, then you're going to lose quickly.

[Elena Baroda]

And are such founders and engineers ample in Europe? Do you have enough of them or is there a shortage of certain types of founders engineers that you like technical people either as founders?

[Edward Keelan]

I don't really have the stats on that, but I haven't necessarily seen that. I wouldn't turn around and go the biggest problem here is that we don't have great universities and we don't have great research. I think where the UK and I'll talk more for the UK are probably less tapped into the kind of wider European ecosystem. But where the UK I think has failed with universities is that again this is probably a whole podcast of thought. So I won't go in too deep. But where we, I think we have failed in the past is we have not been able to take university spin outs and really kind of develop them into super large companies other than in sort of maybe deep technical stuff like biotech. I think what we've seen with this wave of AI is it's really pushing universities to say actually it's not good enough to just have these little incubators where people do some research and we get one or two firms that specialize in university spin outs like the universities. Now the students are coming out there going no, I want to build the next application layer for health tech or for manufacturing or for whatever. And the universities are. I've really seen the step change over the last year. The companies coming out are much more commercially minded with this technology than historically as I said which would. They would be more like deep research based type companies. Universities like Oxford and UCL and Cambridge are just seeing like much more of a. So I think we absolutely have the people. I think what we really need is the Flywheel and by that it's amazing to see how many founders come out of other great companies and not just even come out of, but see visualize other great companies. So if you look at and this is where like Sweden is like absolutely. I mean Spotify I think not necessarily through the Spotify teams but just having Spotify has I think inspired Legora and lovable. It's given those founders. Just because I'm in Sweden doesn't mean to say that I ain't going to build something like Spotify or something like Klarna like really game changing companies and we really need them. We really, really need those kind of foundation businesses. I always sort of talk about, I think Scotland's really interesting. Obviously skyscanner comes out of Scotland. Everything when you talk about Scottish entrepreneurship in tech like often points to the skyscanner kind of like what happened there. Yeah, I really hope that through the success of Revolut and Monzo and other great sort of UK based businesses that we can hold and frame those ones because that's what's going to inspire the next layer as a founders, either coming from those leadership teams or actually just saying, if they can do it, why can't I?

[Elena Baroda]

Where do you see kind of the hubs of really this engineering talent problem? If you want to speak obviously about Europe, feel free to. Or just point to the uk?

[Edward Keelan]

Yeah, it's everywhere. I wouldn't necessarily say some of our companies have got fantastic development teams across Eastern Europe, but that's been a trend for a long time. I think the near shoring obviously with what's going on more geopolitically that has become more difficult. But engineering talent is everywhere. We've got a lot at the moment in Portugal, in Spain, if I think. I wouldn't say there's like, oh my goodness, if you want the great engineers, they're all coming out of whatever. So yeah, I couldn't be specific. And obviously within the UK it's the normal hubs of engineering, the triangle.

[Elena Baroda]

It's quite distributed. Right. When it comes to engineering talent. So it can be anywhere. And would you also say it's possible to build a company from anywhere in the uk, in Europe?

[Edward Keelan]

Yeah.

[Elena Baroda]

Or do you have to go and move to London and you have to move to Paris and you have to move to maybe Berlin?

[Edward Keelan]

Well, I think everyone should always move to London because that's where we are. So please come and join us. Look, can you talk about my experience and how I've invested? So I'm careful to not frame this in a kind of like a researched way. Like from my research I find this. But I'm just talking from like personal experience really, but from my personal experience is that London is a great place to start companies. Not necessarily because that's where the engineering talent is, because I actually think you can have the engineering talent, you can start the companies anywhere but like it or plumb it. London is in the English language and the biggest markets in the world are in the English language. There's not a huge call for people building companies in the UK saying what we really need to do is get ourselves into this French speaking world they want to keep in the English speaking world. And that ease is just that much easier when you're in London. So does London benefit from engineering? No, I wouldn't say so. Not. I don't think it's got particularly better engineering talent. But does it have other tangential things that are really important and why it's such a great place? Yeah, and it's a fantastic stepping stone over to the large US market, which is kind of the mecca of all software companies really. There's a reason why the likes of Ligora are throwing the kitchen sink in 11 labs, whoever it is throwing the kitchen sink going into that large US market, because that's where you build really big businesses.

[Elena Baroda]

When you look at where is the constraint, you said in Europe, the constraint is not necessarily when it comes to technical talent, is there a constraint when it comes to kind of capital density? Is there a constraint when it comes to the speed of iteration in some way or form?

[Edward Keelan]

Might be a bit controversial in this. I don't actually think there is a particular shortage in capital in Europe. Great companies will get funded. Just look at the big companies that have been built like the Lagoras and the Levin Labs, and they're not struggling for funding. So I think the biggest problem within Europe is our ability to take early stage kind of startups that are scaling, keep them in Europe and then build them out bigger and turn them into the next Spotify. And actually I think that is changing. I think companies like Lovable, for example, would have been like in years gone by, may well have been sold like very, very quickly to a big American anthropic or something. And I think it's our ability in Europe to hold on to the ownership of the companies that we can then build them much, much bigger, such that they build those beacons for other entrepreneurs to not sell early to carry on. Clearly there are structural disadvantages in Europe around the fact we don't own large language models, then we don't own the chip manufacturers and we don't do these things. But I don't think necessarily that should prevent us from building some really big companies. So I don't know if that's a good answer, but I don't think it's a capital issue. I think it's a kind of ambition to build really big.

[Elena Baroda]

So you see the constraint with the ambition to build big at times and

[Edward Keelan]

risk appetite and to keep following that risk appetite to keep pushing. Like once you get to a certain size, I think we probably, and again, I don't have the data on it to hand, but I think we probably European founders will tend to exit much quicker than say a U.S. founder.

[Elena Baroda]

And that's a cultural thing. That's a cultural thing.

[Edward Keelan]

I think so.

[Elena Baroda]

Confidence, maybe.

[Edward Keelan]

Maybe. I don't know. But I mean if you just, you know, you look at the big, the big seven, they're all US companies. There's no European companies in that big seven. We haven't built those. And so there is obviously something going on because I don't think it's talent. Maybe it's the way in which we kind of spread around Europe. There is too many constraints, regulations, I don't know. But actually I think there is a bit of a cultural thing going on there as well. And I think that's. And as I said, I really think that's changing. I really feel like we're seeing a step change. I think there's a desire for European founders now to build much bigger, and we're seeing that.

[Sponsor Voice]

TurboPuffer is how companies like Anthropic, Cursor, Notion, Atlassian and Ramp ship their most ambitious search features. TurboPuffer is a serverless vector and full text search engine built on object storage. It's up to 95% cheaper than traditional search databases and just as fast. With TurboPuffer you can index and search 50 million documents at 10 millisecond P90 query latency for less than $100 per month. Head to turbopuffer.com sed to get your first month free.

[Narrator]

You know Fidelity is a financial services leader, but did you know that inside Fidelity is a community of technologists working together to shape the future of finance and tech. Fidelity is always investing in tomorrow. From emerging tech to cutting edge tools that will transform what comes next. Their technologists are encouraged to keep learning so they can expand their skill sets, explore new ground, and stay ahead of this rapidly evolving industry. And right now, Fidelity is hiring technologists to join their team. Fidelity technologists get the best of both worlds. Startup energy that's grounded in the stability of a financial institution. That means support, resources and amazing benefits. Bring your skills to a culture where you're empowered to dream big and build the tech that drives an organization and makes a real impact on people's lives. Find out [e-mail] that's tech.fidelitycareers.com Fidelity is an equal opportunity employer. Every AI team eventually hits the same wall. The models are solid, the infra is solid, but the data coming in is hours old because the pipeline is batch when it should be streaming and nobody's had time to fix it. That's not a modeling problem, that's a pipeline problem. Estuary gives you CDC batch and streaming in one platform. 200 connectors live in hours, not weeks. Your AI is only as good as your pipeline.

[Elena Baroda]

Estuary.dev so that's kind of interesting because it seems that AI is really accelerating. Obviously how fast you can build something, what team size you need to build something in engineering and yeah, that also fuels ambition.

[Edward Keelan]

Absolutely. I've been around for a while because I'm quite old. I mean I was in San Francisco during the first dot com and the old saying that history doesn't repeat necessarily but we can kind of learn from it, it replicates. And so the dot com boom was very, very different to the current AI boom. Because in the dot com boom I was there, there was loads of crazy companies that were never going to like they were never revenue generating and the unit economics made zero sense and all of this good stuff. So when the money stopped, the companies all stopped because they didn't generate any revenue. The AI boom is completely different. It's completely different. These companies are generating revenue really quick. We're investing into companies, you know, likes of Definely and Salesape and Heels. Good and like they are, they're real revenue generation and it is growing rapidly. So it's like these are not companies that are just kind of in hope value, you know, getting eyeballs on things. They're actually generating real tangible revenues. If I look back over the last 10 years of like hype cycles around the metaverse and hype cycles around crypto and blockchain, none of those felt that exciting if I'm honest. It's not like some people thought it was big changing, I wasn't sure solutions, looking for problems. That is not the case with AI. This is an industrial cycle generational that is completely going to change the game I think, and I think we very much believe that because we're seeing it play out just very, very quickly. And I think the people and the founders that are leading into that, that brings huge amounts of opportunity because there's tons of stuff that hasn't been invented yet. You know, say take a blockchain. When blockchain kind of technology, everyone's trying to. I'm going to build a blockchain for, oh, I don't know, supply chain management. And do we need that? I don't know. Maybe we do. It's like with AI, it's like I'm going to completely change the way in which you create presentation materials or I'm going to completely change the way in which you test products. It's like everything is being recreated again. And therefore as a European ecosystem, it's on us to make sure that we're at the forefront of that, that we're absolutely building into it and we're the ones as well that building it. And this is moving so fast by the way, that if we haven't started doing it already, then we're too late. Our biggest fear, probably our biggest fear as a team is like, and we've been investing in AI pretty hard now for a couple of years, but are we too late? Have we missed the boat already? Which might surprise some people, but that's just the kind of way we feel.

[Elena Baroda]

Is that also like, let's say if you are looking at founders or very successful engineering hires at companies, at successful companies that you invest in, do you want people who are AI native? Is that something you really, really expect people to be? And what would that actually mean?

[Edward Keelan]

Yeah, I think so. You know, I can't be like, obviously given the confidentiality, I can't be specific about the companies. But like we've literally removed a CTO from one company who was probably more of the old school, an old school managerial type CTO and moved somebody who is absolutely at the forefront of the world of AI because it's like if that is not where your brain is, that's not what you're thinking about, then you're going to very quickly lose, you're going to be overtaken. I mean the way in which we see the competition of AI, I can go into if you want, but like if you're an existing software company and you are not pivoting and turning and thinking about this, you will be destroyed quite quickly one way or another.

[Elena Baroda]

So there has been what public markets called Softmageddon, which basically meant that there was like many public companies whose stocks crashed because of, you know, court with new tools. Tools came out and we can also go into the details. So that actually happened to mandate. So what happened is that was basically back in February where it's like ancient history.

[Edward Keelan]

It happened two weeks ago and it's like everyone's now like, oh, that's two weeks ago. What's happening today? What's that?

[Elena Baroda]

That was really, really interesting. That really spooked kind of also this AI eating software narrative. And separately there was a CNBC reporters with zero coding experience and used Claude code to build a working Monday.com clone in under an hour for very little money in compute credits. It was like five to fifteen dollars. And that went viral obviously and hammered the stock even further. Yeah, so it's kind of interesting to look at that and is actually AI eating software and how does it affect your investment thesis?

[Edward Keelan]

Yeah, I mean it is literally our conversation. Probably every other hour of the day we flip into this conversation and trying to work out what software is safe out there is the big topic of the day. When we're working at stuff. I mean my hypothesis changes on a daily basis about this, so I'll be careful. But if it's okay, I'll sort of double click slightly into this. It might take a couple of minutes to sort of explain my thinking. But I think first of all you can sort of split the world into things that were built pre AI pre the big 22 chatgpt bang and AI native software. I think if we're looking at AI native software and generally when we're thinking about new investment, we're talking about AI native software. That really the big question for me there is can you replicate that software using an LLM? And that's Anthropic. Go hey, we've got a new legal plugin all of a sudden, Legal tech, even native AI legal tech. Unless it has a very specific purpose and we're invested in definely a legal tech tool which we believe does have that very specific purpose that can't be replicated by LLMs but the generic ones. There's going to be a huge amount of pressure there. Why do I need to buy Harvey if I can just get Anthropic's plugin to do this at a tiny fraction of the price? And I that to the kind of the spellchecker world of software. Like back in the 80s it was like 35 spell checking companies and then Word just put a spell checker plugin in and all the spell checking companies died. So the same sort of thing could repeat there. But AI native companies that have a very specific purpose that aren't going to be disrupted by the LLMs, that makes much more sense. That's kind of like where we're thinking we're going to invest into. So we just did invest into Heels Good, for example, which is using AI to disintermediate agency from healthcare. It's like Anthropic is not likely to build the application layer to disintermediate healthcare resourcing. It's just too verticalized for them. That's AI native. If you then think about the older softwares, the Mondays.com I think they fall into some broad buckets of companies where you go are you kind of an old school creation tool? Are you building stuff on a platform manually? They're dead. I mean creation tools are dead. Lovable replit. They're building products, AI native tools like that, they're just building product really quickly. And I think those creation tools are going to really struggle if they can't switch and all of this is if they cannot, then switch themselves over and they can't tilt into that kind of AI world. You've then got like discover what I'd call discovery tools. So things like data analytics tools or penetration testing tools and things like that, I think that all becomes about the data. Like if you're a discovery tool that really owns its own data and has that flywheel, then you may be safe. But if you're just picking up third party data from here, there and everywhere and then presenting it to people, or you're that type of thing, then I think you're in trouble because it's like AI agents can go and get that data and they can present it much quicker than you. And then the sort of the final category and the one that's more complex would be around system of record and workflow, and then with system of record and workflow. So I'm talking about your kind of vertical SaaS, ERP systems, your CRMs, those types of products, your accounting software or whatever it might be. It's like, first of all, can your workflow tool be easily built by AI? You might say this is like a Monday.com problem. And it's like, yes, it can. Like AI can just build what you've got. You're stuffed. You're going to be in trouble because companies will probably build a lot of that type of basic workflow stuff themselves using AI, because why wouldn't you? It's free, you don't have to have headcount. And then you get onto like if you're a system of record, will there be naturally an AI native component competitor? And if you're a workflow management tool or a system of record, where actually right now there isn't one, there isn't any AI native competitor, Maybe it's an accounting solution or something like that. You're probably all right for a time.

[Elena Baroda]

You recently said growth at all cost is out. Metrics like profitability burn, multiples, gross margins and rule of 40 now matter immensely. Can you explain what you mean and what is the change for who gets funded for engineers? Is that again driven by AI, that it's really about growth at all costs being out?

[Edward Keelan]

I don't know when I said that, it's probably like a few weeks ago, which makes it, as we've talked about, immediately out of date right now what we're getting. We are in a massively polarized world. So you've got AI native businesses that are growing at all costs still, and nobody cares about buying multiples. No one cares about that stuff. It's like if you're growing super fast, and the funny one is no one cares about gross margin anymore. I listen to podcasts and listen to hear about businesses that are growing super quick. They're the darlings of European tech, I won't name them. And then they're like, oh, and what's your gross margin? And they're like, well, I'm happy to tell you my top line ARR metrics, but I'm not going to tell you what my gross margin is. It's like people just don't care. It's like at some point we'll work out the token costs and whatever. But that kind of scares me. I'm probably not brave enough as an investor to just dismiss the fact that you're losing money every time you win a customer. So, yeah, you've got this world where these types of companies over here will get funded on the basis of their growth. And the burn and all the rest of it is just not important. On the other side is if you're not kind of hypergrowth, if you're not super growth, then you are still in a world where metrics really matter. So if you are a kind of vertical SaaS solution that's leaned slightly into AI, you've got a bit of an AI story, probably still fundable. But if you're growing at 100% a year, say, take a company going from a million to 2 million arrows annual recurring revenues. Sorry, so talking financial terms, yeah, £1 million worth of annual recurring revenues, £2 million worth of annual recurring revenue, and you've burned £10 million to get there. So you're at what they call a 10x burn multiple. You're not getting funded, you're not getting funded, you're not growing quick enough. As an AI native company, you're still in that kind of vertical SaaS world and therefore you need to basically have a burn multiple of like one or maybe one and a half. So you're basically saying if you're adding a million pounds worth of aro, the most you can really afford to burn is a million pounds at this stage, you might get away with that. And even that's high. I think just. And I don't know if going off on a tangent here, but one thing that's quite interesting about the whole death of software and suss top licks or whatever they called it, there's so many factors that are coming into this. So retention for a lot of these products is still very high and it will remain high for a while, but seek count is going down because people are hiring less people and therefore they require less seats. So on one, there's like revenue pressures coming down on these companies as seat count comes down and growth is harder. But on the other side as well, you see with block, I mean, people are firing people. The numbers of people that are required to be in these companies from sales and marketing and data analyst teams and all this are coming down as well. So theoretically, EBITDA margins should also be coming up. And so when you're looking at the kind of earlier stage versions of these companies, you're like, actually, this is not 2020. It's no good to just hire endless engineers anymore. Why are you hiring all these engineers? What is wrong with your code? Because can't you just create it using any of the many AI code generation tools? You should be much more efficient. So we are expecting companies as well, just to be efficient, not only because burn matters, but also because the tools are there to be more efficient. And if you aren't, then that's a question for management. It's like, what are all these people doing?

[Elena Baroda]

That's obviously something many corporates are starting to ask as a question, right?

[Edward Keelan]

Yeah, definitely. And the people that are getting really impacted first are the new hires and usually the young people. I'm obviously slightly at the forefront of this movement and at the way in which the technology is coming in, I am very, very concerned from a social point of view is like, graduates are struggling to get jobs. Like, the number of jobs for graduates has gone down by 25% over the last 12 months. That number is just going to increase because I think companies, large companies are going, we don't want to hire, we don't want to hire graduates. Maybe not, because today we're ready with all of our AI tooling, but in 12 months, 24 months, we need to be more ready with our AI tooling. And if we're not, I know that our competitors will be. And therefore, if you're at the top of an organization making a strategic choice about whether to hire your next batch of graduate engineers, you might be like, well, we'll just part that decision for a while and see if this AI tooling can do it for us. Because what we don't want to do is hire them all, build up that cost base, and then have to somehow row back from it. So, yeah, large companies are absolutely making those decisions. But at the moment, I think it comes at the entry level, and I think that's really hard. I don't know, how do we describe that to our Young people and people who have spent the last 10 years learning how to be a computer scientist. Some of them, the top ones, the ones that are at the absolute forefront, they're going to get paid megabucks. Those ones are going to zoom into the best positions because it's such a race for AI talent, But the bottom 50%, that's going to be much harder.

[Elena Baroda]

Where is Europe in the AI race? So we are always talking about us versus China at the, let's say, kind of layer where we have lovable, we have Ligora and companies like that. Can Europe win? And you also said the talent is here, the capital is also available. Can actually Europe win? And what does winning actually mean?

[Edward Keelan]

Yeah, I was actually going to push that question back to you. It's like, what does win mean? What do we want to do by meaning win? I don't think society necessarily wins because. Does American society win? Because Elon Musk is a trillionaire? Because they put these massive. I don't think that's great for society. I think there's disparity between. Wealth is grown tenfold plus over the last decade. I mean, it's crazy. No one's talking about it, but I don't think America is winning for a wider society if the winning is just building the big companies. I think what's important is we don't get left behind and that we build our own great companies and we do that in such a way that we lean in to what we're good at. And I do think the geopolitical world means that actually Europe has an opportunity. If we can get together, we can work together, we can share talent, we can share resource. Obviously I'm sat in the UK here, so people might be throwing rocks at me, but if we can do that, we have a real opportunity. Because I think there's a category of talent that would prefer to be here. I mean, I don't know if you're an engineering talent. Do you want to go and work in the US right now? I'd be really careful, probably.

[Elena Baroda]

Well, it really depends where. Right. If it was in New York or San Francisco. Potentially. Yes. Because still, it's just so incredibly interesting to spend time there sometimes. Right. Just to see the raw ambition people have.

[Edward Keelan]

Yeah, I mean, I love New York and San Francisco, having lived in San Francisco. So I'm just meaning more like I think there is a moment in time for Europe right now and I'm seeing us start to capture it. I really am. I truly think there has been a step change project Europe and I say the way which the universities are starting to actually kind of get their act together, I just, I think there is a moment in time where we can say actually, and I would say Sweden has done this better than anyone right now. I mean the evidence is that culturally they seem to have absolutely got it, they seem to be doing it. But yeah, we've got that moment in time right now and I really hope we lean into it. And I think we will. I actually genuinely think we will. I think it's going to be a really exciting time. And the question I flip around on is whether Europe should really get their act together and build its own super large language model. Should we be doing that? I don't know. You're probably better positioned than I am to answer that question. But just almost more as a defensive move. But then. And building technology for defensive moves isn't probably the right thing to do either. So who knows?

[Elena Baroda]

We just talked about Europe also building great companies and part of that is also having just the environment to do so, the political and the business environment to do so. And this year von der Leyen, who is the president of the European Commission, she just announced EU Inc. Quite interestingly that was initiated by the ex CTO of Product Hunt, Andreas Klinger, he's also European person, has been campaigning for this really like a champion. So what it promises is a 48 hour incorporation across the whole block and standard ESOPs and it just costs like a euro which is a bit over a dollar. It would come in effect sometime in 2027. So it's definitely still more than a year out till it becomes effective. Like do you think that can actually change the engineering reality for people, the founding reality for people in Europe? Would it be relevant for the company or solve problems for the companies you invest in?

[Edward Keelan]

Yeah, I mean absolutely 100% the UK. One of the reasons the UK does really well at founding companies is because we have company law. I think our company law is pretty damn good. Like model articles. You can found a company in the UK within an hour. I mean like literally just go online. It's even better. You can start in a company in the UK in an hour with a bank account. Tangentially, my children run Happy fidgets shop which is not currently live and we've got our little company set up and it's like because they're really into their fidgets and don't ask, let's hold on the discussion. But the point being is to set up that company was literally open a Tide bank account, register a Company at the same time, you can start trading, you can start trading on the same day without going into which European countries. Because I don't want to diss on any specific European country. It can be a nightmare, an absolute nightmare. It's like costing hundreds of pounds, some of them. You have to put tens of thousands of pounds into bank accounts. I mean, that is nuts. Say you get a university graduate that's come out of university, wants to start their own company. They're like, right, start your own company to put £20,000 into a bank account. And by the way, you also need to go along physically and you need to get a notary to sit there and blah, blah, blah, blah. And then the company law is horrible. Every time you want to make changes, it's horrible. And yet we've really felt it across Europe where there are jurisdictions across Europe where we have really felt like, wow, not just on the founding of companies, but also on then when stuff happens, maybe especially when things go wrong. God, this is like treacle, like, God, I am, I won't name the country, but it's like we are never doing business in that country again because when things go wrong, it's just so archaic. The shareholder rights that they've got, and not only that, the way in which you have to sometimes have to. To physically be there with all the shareholders to make. So, yeah, anything we can do here to make it so much easier to incorporate and unify companies would make a huge difference across Europe. I think we're quite lucky in the UK that we kind of already have that.

[Elena Baroda]

Super interesting. Yeah, amazing. Which constraints do you see beyond that that need to be fixed to really help Europe become really attractive for founders and really easy to scale companies too.

[Edward Keelan]

So, yeah, beyond the founding and I think we've talked about it in large part. I don't know, how do I put this? I don't think we need to attract founders. There are loads of potentially incredible founders within Europe. So it's not like we're trying to attract the best. What we really need is we need to attract the best talent from China to come and start businesses in Europe. No, no, it's here, it's here, it's ready, it's ready to go. I think what we need to do is build the institutions that encourage it and the universities, as we've talked about at length, and the organizations that really encourage those founders. I think there was. Somebody was talking about something they're doing in France, which I think is really interesting. For example, if somebody wants to start a company, they will pay. And I don't know the rules, so I'm sort of picking up little bits. But in France, they're like, they are paying for accommodation and food costs of potential founders. Who wants to start companies as part of a program to say, look, we know that actually it's really risky starting a company. You know, there's a lot of risk. You put your life out there, you're going to spend huge amounts, hundreds and hundreds of hours, you're not going to get paid very well. The gold at the end of the rainbow comes years and years down the line. It's really a passion that you must have. And so for a lot of people, they just don't want to take the risk. You know, it's like, I need to live. And so I think this idea that they've got in France of actually, I can't remember the name of the program, I'm sure you could AI it and find out what it is, but is around like covering a lot of those living costs to make founding companies feel much more secure. And I think once we start, start to do that, then I think it becomes much easier. So. And I do think it's a flywheel, I think, as hopefully some of these companies do start to get big. You know, I imagine Lovable is going to spout an amazing amount of new entrepreneurs like just like Fuse Energy started out of Revolut, that's an interesting company and many, many others. So, yeah, I think we have to make entrepreneurship easy for people from all backgrounds.

[Elena Baroda]

Does regulation slow down entrepreneurship in Europe? Does it slow down like engineering in Europe? You have to think of gdpr. There are many compliance issues that you might have with enterprise. There are many consumer rights that you might need to take into account when building. So does it slow down?

[Edward Keelan]

Not that I've seen. To be honest. I wouldn't say that I've ever had a company that has said, oh, the biggest problem is the regulations here. A lot of these regulations like GDPR are there to protect customers. They're not bad regulations. They're there for a reason. I haven't seen it in terms of engineering and. Sorry, I'm going to answer a different question. That's all right. In terms of engineering, the biggest problem we have when we invest into a company is around engineering, is where the engineering structures are potentially wrong and where a lot of the legacy ways of doing things create problems in the future. If I looked at, okay, what's been the biggest engineering issues that I've had in my companies versus regulation of that the internal issues with the way which the technology has been developed are way greater than, oh, we didn't comply to the right GDPR legislation or we had this security issue, we had to report this. It's just like you can hire people generally to do that stuff anyway. But I've had much more growing pains with engineering around things like monolithic code bases that have tons. The amount of companies you invest into, you invest in them. Before you know it, you're spending two years refactoring everything. Because actually the layers and layers of terrible coding on top of each other just mean that it's much more difficult. And the big crisis we've had with companies like I've had multiple exit processes fail, for example, because when you get under the bonnet, maybe that SaaS product is all stitched together and it kind of works. But actually people are like, no, that COD is just not. There's too many risks within that code and how it's put together that I want it. So yeah.

[Elena Baroda]

Is that something you try to because of the experience you've had like with as you said, exits failing because of that? Is that something you factor in earlier that you really double check like the technical talent that is hired or that is senior in a growing company?

[Edward Keelan]

I would think as a team we've probably learned that. I think as a team we've probably learned to double down on it. If I had to frame it, it was like technical due diligence used to be a bit of a tick box exercise. It's like, has the technical DD person looked at this and said, yes, that's a technology product, if they have ticket. And now it's like, what's actually coming out of those reports? What are the actual risks and how real are they? And have we seen this pattern of those risks before? Can we start to understand that? So yeah, I would say we've matured as a team to double down on the quality of engineering. And a lot of this happened during 2122 when people just could not hire enough engineers. Like we'd be going around and actually, I hope I don't speak out of town, but I think what happened is a lot of technical leaders just didn't have much experience because CTOs were getting paid more than the CEOs and everybody that has a senior engineer for a year became a CTO the following year. And we just saw a lot of, I think product probably not necessarily built in the best way, maybe as well. And also a lot of money spent probably not building the greatest product Back in that little crazy period where everyone had hiring engineer. So yeah, I think because of that, certainly over the last five years, we spend a lot of time thinking about the scalability of the software and the history of it. Has it been put together in such a way that can continue to scale, that we're not just going to spend the next two years refactoring something and that's not where we want to invest. We're not trying to invest in rebuilding products that are already built, we're trying to invest the growth of those products.

[Elena Baroda]

So following up on that is technical debt. So basically the cost you pay for taking shortcuts, is it perceived differently in European companies than it is perceived elsewhere or in the US specifically?

[Edward Keelan]

Yeah, I wouldn't have a good answer on that. Maybe it's a maturity of market thing. I don't know, maybe the US VC market, investor market has probably had a longer history. Maybe therefore they have a better understanding of it. I couldn't say. But certainly for us now, technical debt is probably a much bigger red flag than before. Interestingly, in the current wave, there's currently a wave over the last few years. It's like the technical debt isn't such an issue for a lot of the companies we're backing right now because they've literally built it over the last six months to 12 months. So when you're investing in a vertical SaaS solution that's been around for seven, eight years, that's very different to a native AI company that's been around for six 12 months. There you have very different technical questions that then becomes about scalability and risks of the models that companies are using, those questions are much more prevalent. And actually the discussions with the technical leaders, the cto, the heads of engineering, to try and understand where their mindset is.

[Elena Baroda]

How does founder led go to market intersect with engineering decisions? So let's say you still have founder led growth and then the founder is kind of like in the position of being the product owner and being the salesperson at the same time. And that can kind of create quite a lot of of pressure on engineering to go and build something even if it's maybe not 100% validated.

[Edward Keelan]

So yeah, yeah, I've written a lot about founder led sales to team led sales and product market. Something I've classed as like losing product market fit. Happens a lot in founder led sales where product teams built great product, founders selling it and then because they're doing well, you're a health tech company and then all of a sudden a Financial big bank comes up and says, I know you do healthcare, but we'd love to take your solution and put it into the finance market. And the founder's like sheesh, that's another half a million pounds. And the product team are like, well hang on a sec. Everything that we've built has been directed at healthcare professionals, not bankers. And you can very, very quickly lose product market fit. And it's like a very disciplined CEO that stays on course and the product teams that go with it. And as I said, so I think a lot about that. Like I am very, very nervous about companies that are trying to boil the ocean and serve many, many sectors and geographies early on because I think that what happens is you end up with what I'd call a Frankenstein product. You'd end up with a bit of product over here, a bit of product over there and rather than having something scalable, you end up with something that actually now you need umpteen different people to maintain various bits for various customers. And that is a nightmare place to end up as a company, as a tech company. And as I said there's this pendulum that goes on where you get like back in the dot com era it was all about engineers because to build websites you needed super duper technical expertise. And then by I just read a book from Julie Wainwright about the real real and she was like she [e-mail] in 2000 it was all about engineering. By 2018 when she started the real real it was all about actually engineering wasn't that important because anyone could build a website. It was all about distribution and blah blah, blah. I think that pendulum then now has swung again because you're now in the points like it's the engineers and it's the product people that know what is possible and what solutions to problems are doable and how it can be used and how it can be app for the next five years. I think engineering is going to have a huge role, huge, huge role in actually how the go to market function works within companies because it's only the engineers that are going to really understand what is possible. Which is why I say we probably over index on that as we're now looking at potential future investments. Difficult for founders, difficult to founders to how that works.

[Elena Baroda]

Edward, if you had a friend who had a choice to make and an engineering friend and he or she is choosing, or they are choosing between a well funded European series B and a US offer with a higher salary, what might they not be thinking about that they should be thinking about when making that decision.

[Edward Keelan]

Oh, I mean, I would have to get to know the friend better first. There's a lot in that, and I think it depends so much on their personality of who they are and what they're looking for. Like the US world of startups, as you say, is unbelievably exciting. It's unbelievably brutal. Like, you go over to the States, you start in San Francisco, join a startup, six months later they've run out of funding and you've got two weeks notice and see you later. You'd think culturally, the European startup is probably much more secure in terms of job security. I would like to hope that more and more engineers will start staying in Europe and will want to build in Europe and will want to see us build the next great seven. Maybe not the next great seven companies. Maybe that's a silly ideal, but the great companies that will continue to spawn other great companies again. Again, that is my hope and I truly believe that's going to happen. So I would encourage them to stay in Europe for that reason.

[Elena Baroda]

Yeah. Thank you for the conversation, Edward.