OpenAI's Laura Modiano on hackathons, Lovable and how startups can partner with the tech giant

0:02Hello and welcome back to another Sifted podcast. This week you're in for a treat. I have someone sitting here with me today in the studio who has, within a relatively short space of time, come to play an incredibly unique and valuable role in Europe's startup ecosystem. She is OpenAI's Laura Modianu. Laura, who leads the tech giant's startup team in Europe, is seemingly forever on the road at a hackathon or on a stage with Europe's leading AI companies and investors. She's also a very active angel investor, networker, and, you may not know this, a huge puzzler, jigsaws and crosswords. Today, we'll get into what exactly it is Laura and her team does, where Europe's AI ecosystem is heading, and what she's learned from some of the world's best founders.

0:47Laura, welcome to the show. It is such a pleasure to be here. Thank you so much for joining us. So you joined OpenAI a year and a half ago, tasked with supporting startups in Europe. Can you briefly explain how your team actually does that? Yes. First of all, I always say this, my job is a real privilege. It is an honor and a joy to see the European startup ecosystem grow, evolve, and really become on par with any other ecosystem in the world. How we work is we try to be very present in the ecosystem and to partner with startups. not just support them, but truly partner with them on a couple of things.

1:36One, to understand what they're building, how they're building it, how they're using the technology, and what it is that they want from us. Because at the end of the day, our job is 50 % to represent OpenAI here and to the ecosystem. And the other, a lot more important, arguably, 50 % is to represent the voice of the builders in EMEA back to OpenAI. So what we really care about is to understand how they're using the technology, what is working, what is not working, where we can improve, how the feedback can lead back to model improvements so that we truly represent their voice, so that we're developing and giving the tools for them to build the use cases and serve their customers.

2:20So that is our job number one. How we do it is we have teams that work directly with the startups and then AI engineers and applied AI engineers that work with the teams in order to really help and unblock things, accelerate the building. Then we also work with investors to make sure that we're covering the full portfolio. So as soon as a company gets funding, that we're there to support them from day one, that they can start with us as a partner, use us however they want. And then third, to just be in the ecosystem and meet developers sometimes even before they found. So whether it is at universities or hackathons, to just really give people access to our tools and to get hands-on experience with us.

3:02That's why we also have a fantastic developer experience team. We do a lot of meetups, a lot of builder workshops, and then we talk a lot about codecs. And I mentioned at the top you always seem to be at hackathons. I'm guessing that's because that's one of the best sort of touch points or ways of understanding new ways people are using the technologies. Is that right? I love hackathons because for a couple of reasons. One, this is a really incredible moment, not in one market, but in every single country that I go to in Europe. And European builders really have true ambition and excitement about technology right now.

3:42I think what has changed from hackathons sort of a year ago to now is that in the past, we used to see a lot of people going to hackathons to try technology, build a cool demo, and then go back to their sort of day jobs. Now we're seeing people who are really sinking their teeth in technology and going so fast from idea to MVP to product that is ready to commercialize. So what they do in hackathons is accelerate existing workflows and they use those to either give us feedback or to see how technology can move a lot further to accelerate what they're actually working on. And for me, the reason why I go is because I want to be where the builders are.

4:21I don't want to be far away. We're not a faceless company where we want to be close to the people, talk to them, understand what they're doing, see how they're working and be available for conversation, feedback and actual exchange. And then I often follow up with them and see, you know, if something they said that either wasn't working or was working really well, that we can follow up and actually have a conversation about it. What are some good examples of things that have come about as a result of you being super close to those startups or those builders early days? Like what are some things you've actually fed back and have, you know, led to product changes or whatever it is?

5:02I think, I mean, there are things that happened literally a couple of days ago. But for example, we're now looking to specifically understand how builders are using codecs in their workflows, what it is that they're building and how so that we can support them, but also understand how a product that we're iterating continuously can improve. So I was just in Switzerland last week and I had this fantastic breakfast with I think it was about 50 or 60 builders. And they were building everything from Vibeflow, which is a company that came out of YC that started in Zurich, all the way to humanoid robots that can be deployed in space.

5:43And I was hearing about how Codex has accelerated the web app and the control center that can control the humanoid robots all the way to how there was a woman who used to work in a big tech company that then very quickly, because of a hackathon that she was in, was able to build a company that started having some initial revenue for her, which then gave her the opportunity and the tools that she needed to say, I can actually build what I started in a hackathon into a real company and I can go out to fundraise. The other thing is that often I meet founders or aspiring founders who are looking to meet more investors.

6:23And because of the work that we do, we're connected so much with the ecosystem that when I see really interesting founders, I try to connect them either for advice and mentorship or for fundraising when they're ready to fundraise. And how do you pick who you partner with? Because I'm sure every AI company in Europe wants to run a hackathon with OpenAI, do something with OpenAI. How do you select both the investors and the companies that you spend more time with? So we partner with everyone. It's not that we have a preference here or there. We're here to work with the ecosystem as an ecosystem as a whole.

6:59From the investor side, we work with all the funds. any fund that wants to work with us can ask us and if they don't already which most of them do that can be part of the open AI for startups program they can offer credits they can offer access to chat GPT and codecs for business so we're here to support everyone without any differentiation for events I think it's more people who are actually asking us so there's a level of proactive so for example if a fund says we're thinking of doing event xyz whether it's a hackathon, a dinner or a meetup, then and they want to partner with us, we're very happy to do that.

7:36Again, we want to be where the builders are. So, for example, last week or the week before, I was in Berlin for an Open Claw 0.9, an OpenAI and PostHog actually event. And then I was having conversation with Christoph Jans and the team and they said, could you please support? Are you interested? And we were extremely happy. And actually, it was such a fantastic event because, again, I spent the whole day there. Everyone was very technical, very open, had specific feedback, was building in real workflows, was using codecs across the board. And then there were some fantastic demos of how people are using codecs.

8:14And there were people saying, oh, I have, you know, five terminals open at the same time. And this wasn't possible before, which is what we love to see. So the accelerated technical skill is what we're usually looking for. And for someone who is listening to this, who's maybe never been to a hackathon, is a bit sceptical, thinks they're still how they were five years ago or whatever, or thinks they might just be quite for, I don't know, very young men or bros. What would you say to that? You have probably not been to very many recently because it is such a joy to see how many technical women there are.

8:51Maybe not every single hackathon will have the same split, but more and more, we are seeing more women who are very technical, who are extremely capable. And many of them, we see the women actually winning the hackathon. So this happens to be in the US, but a month ago or so, I was in Stanford for a global hackathon that was happening at the same time in Stanford, in MIT, in Zurich, in London, in Munich. And it was two women. who won and they were exceptional. And the whole event was organized by a woman. And actually, I was sitting next to them and seeing them build and they were phenomenal throughout the whole time that they were building.

9:32In Zurich, as I was just saying, I saw some fantastic women as well who are building in robotics, who are building in vibe coding, who are building in agentic capabilities. So I don't think that the perception of how many men and women are there should stop you. If you're a woman, you want to participate, go for it. And you're the one who's going to drive the change. So we just have to show up. And as a woman, I have become a lot more technical because of tools like Codex and other tools out there, such as, you know, Lovable earlier on. And I love building. It is also, hackathons now are taking a lot shorter time.

10:11So whereas before you might have had to go there for a whole weekend, there are many events that are three hours, six hours, one day so if it doesn't take a huge time out of your schedule even as a working mom I can do it so any other woman can do it too. And on the topic of bros we've recently been writing about how it feels like in tech the tech world bro culture feels like it's made a bit of a comeback it feels like there's lots of men who are very vocal on LinkedIn who are talking about these kind of work practices like 996 or whatever that are to you know broadly maybe less appealing to women or less uh you know inclusive to lots of different types of people do you think bro culture is back I think hard working culture is back and I think working very long hours is definitely back but part of why this is back I think is because people are just really excited with the new tools and it's they're they move so fast and technology is moving so fast that when you get excited like I started building for myself I started building for myself at work and for myself in my private life and honestly all weekend I spent just building my open claw and her name is Demi and she's very cool and I was doing it with my family and did I end up spending you know 18 hours a day on AI for fun Yes.

11:39Did I have a great time in it? Absolutely. Did I work really hard? Also, yes. But I think at a time where technology is moving so fast and it's so much of a differentiator, the speed with which you move, it is important to be a hard worker because it is speed that sometimes can be emote. It is also taste that can be a mode. And what I think is a unique differentiator that we as women have is that we often have a different point of view and wonderful taste. And when taste is what can be your advantage, I would love to see women's point of views and tastes be represented in the technology that is being developed now.

12:20You mentioned to me before we came on air that you've also noticed a bunch of hacker houses and different communities springing up that are more geared towards female engineers and developers. Tell us a bit about those. Yeah. So there are, first of all, it's great that across Europe, we're seeing more hacker houses. So there's Talos House here in the UK. There's Freight in Helsinki. There's a base space, I think, in Lithuania, if I'm not mistaken. There's multiple places and it's fantastic. and what I've been hearing recently is that there's a couple of those that are actually so her circle is in Switzerland and I believe tech bros is actually one called um here in the UK which is saying if it's a house where you're living and co-living and co-working and you're there 24 7 then if you don't want to be stuck in a house with men all day you also have an option for women I'm not saying that one is better than the other but I think it's great that there are many different spaces and any individual, no matter what their sexuality is, can choose to go to the one that is right for them.

13:25And these communities are there to be inclusive and bring more people in tech and have more founders that are emerging out of Europe. So we've got hacker houses, we've got hackathons. What are some of the other best sort of touch points would you say for either people listening to be involved in or that you go to and you find most sort of valuable? Yeah, I think a lot of meetups. So, for example, AI Tinkerers, I think, is doing really good meetups and a lot of the conferences around as well. So you and I have been in many conferences together. I think the Sifted Summit in October is great, where it's not just people who are building together, but also hearing from people who have done it many times before.

14:04Because I think what Europe really needs is the community and is the North Star of this is possible and this is locally possible. We have incredible founders in Europe who have done it not once, but some of them multiple times and who are really paying it forward. And the co-location, the idea sharing, the building in public, the building in community, I think is what the biggest accelerant to our ecosystem is going to be, as well as bringing people across communities together. So what I'm seeing now, which is, for example, hackathons used to happen in different places, And now we're seeing some of them take the winners across each country and build them together in like a super hack.

14:47So the more you can connect with others and get feedback, get recommendations, get pushback, not building to 100 percent, but building to 80 percent, showing it in public and then finishing it up with the community and the feedback, I think is where we're going to get to move a lot faster. Interesting. And when it comes to when you actually kind of go and work with companies, I know I've seen your LinkedIn post that you spend quite a lot of time in lovable's office, for example. What would you see a kind of key attributes or working patterns or ways those company CEOs are like? Can you distill something?

15:30Yes, I think the most successful companies have two things that they do extremely well. They know how to learn fast and they know how to unlearn fast. So they understand that playbooks are not static. No matter how much growth you have, no matter how fast you're moving, you have to continue to be driven by your users and by the product that you're building for your users. The same way that for us, we have products that we want to listen to our users and we want to continue to focus on what it is that they want and how it is that they're using our products. I think a startup should always have that driving force.

16:10And if a playbook worked for your zero to one, it might not work from the one to 10 and it might not work from 10 to 100. It also means that the team that you started with and how they used to work might have to evolve. And if you bring people from the outside, they might come with some experiences, but the playbook has to be redefined for the company you're building now. So talking about Lovable specifically, they have an incredible team that reinvents their playbooks constantly because A, they're helping recreate a new category, but also evolve it as their users evolve as well. And I think that openness to continue to learn and to continue to develop the playbook, I think is the one thing that will change the speed and the pace with which you're going to grow.

16:55And are there any sort of habits, rituals, weekly, I don't know, meetings or things that you see, have seen companies do that you think that's really cool? Yeah. That's really helping them achieve that learning and unlearning process? A few things. First of all, I think it's companies that are, actually, if I take a step back, one is because we're a multicultural region, but we want to be global. It's understanding that you can be global from day one, and if not truly global, then multinational. So taking away any preconceptions that you might have of this is how a market works because I'm only looking at my local market.

17:35So one is really trying to have a simplification of things. And the more you pare it down, the more you go back to the roots of the root cause analysis, then you can take out what some of your assumptions of what a workflow look like and say, how do I bring value to its core? That's the first thing. The second thing is everyone in the team has a voice and usually it's more of an equal voice than I think traditionally we would have seen. So now we're seeing, especially because tools in AI, ones that we bring to market, have given so much speed and power to everyone in the team. A lot of the teams are a lot smaller.

18:13So you're seeing the marketing person, the product manager, the engineer, the comms, the devs, all of them have to both engage with a product, with a process of building, but also with the customers. And then they all have to be building in public again. So I think we've seen so many posts about this that everyone in the team now gets to have a voice on social media, gets to say how they're building and what they're building, why it is that they're building and for who, and getting that feedback continuously in, but sharing it back in and saying, you know, marketing person, PM, have a conversation and then have a say as to what is happening in the product and iterate based on that feedback as a team.

18:55Where are you seeing the kind of very, very fast scaling companies struggle as they hit scale? Like what are the points that are proving most difficult? I don't think that there's one rule for that. I think different industry companies will have different things. So for example, if you're building in a regulated market, it might be regulation. If you're building in an industry where data has to be collected in a very different process, for example, in healthcare data, because the healthcare systems are different, that might be the scale with which you go to different markets. But I think at the end of the day, the ones who grow really fast at some point decide whether they want to hire externally someone with a lot more experience.

19:38And you go back to that works if the unlearning and relearning and building the company culture happens well. And then always, I think, no matter the size, how do you maintain culture and a culture that is true to this company and its DNA is one that you have to work on on a continuous basis. Who or which is your favorite company to partner with? You have to tell me one. I mean, right now I do work a lot with Lovable and part of why I love working with them is A, they're deeply technical, but be across the team everyone again has a say so whether i'm talking to their community team or their devs or their engineers or anton the founder himself everyone has actual feedback real passion for what they do they work closely with us and it's a live conversation that we have on an ongoing basis and when it comes to a company say like lovable partnering with a big tech company like open ai what what do you think is the best way for that to work because i sometimes have conversations with people or with with investors or with founders who sort of say they're not sure how how close exactly to get to you guys because after all you have so much money and and so many brains in the company and there's a slight perhaps risk perceived risk that if if a company gets too close to you then you you know what they're doing so well you could do it?

21:05We have so much passion for startups and a very large team whose job is only to make startups successful. It is my job. So hopefully that's what we do because truly what we want is to make startups successful. How we help them is we, especially companies like that, that are very heavy coding generation users. We work with them on the applied eval side to see within their use cases, how our models perform, where and how we can help them, whether it is because of how they're prompting, because of how they're using a model, maybe because a new model came out, how to then make sure that that model is optimized for their use case.

21:50And again, our applied AI team works very closely with them. And it seems to be a very productive partnership that we have. And to get kind of very, very blunt, like what's in it for OpenAI? Is it that they will use your models forever and not another big tech companies? I mean, hopefully. And is there any financial, you know, do companies reach a point where they're paying you for some of these advisory services? No, not at all. What they pay for is the model usage. So tokens and that's it. And we're very happy to have a very partnered model where we make sure that our models actually work for them.

22:32So as long as our technology, like we're a research and deployment company. We're a company that creates models for usage. And we want to see people build companies that are successful, that are creating value, that are creating opportunity. And that is what our customers across all industries are doing. And you mentioned quite a bit earlier on that you're not just working with companies that already exist, but you're also trying to find people who are still at university or maybe in research labs. Which are some that you find incredibly impressive around Europe? So ETH Zurich, I think is fantastic.

23:09KTH in Sweden, as well as SSC, so the School of Economics, Stockholm School of Economics. Obviously, I have to say Oxford, Cambridge, Imperial here in the UK are excellent. I was at the Imperial AI Society a couple of weeks ago with Peter Steinberger, the founder of OpenClaw, who now works at OpenAI. And I was very impressed by the students. But equally, you see universities across all of Europe. So whether you're in France or in Amsterdam or the Netherlands or in Germany, you know, CDTM is one of the programs in Germany that has had one of the highest success rates of founders coming out straight from university.

23:52And I believe they've had nine unicorns come out just of the CDTM program. So we try to work closely with them from day one so that when they start the company, we're already there. We're accessible. We're there to help. We give them credits. We have technical teams and business teams who can work with them. And we're there to partner for their long term success. What would you say the special source, if there is one, of those institutions that you just mentioned that are doing it really well or sort of instilling entrepreneurialism and enabling people to actually spin out well? Yeah, I think it's mixing theory and education with actual entrepreneurship.

24:35So being very, very technical works if you're taking that technology and commercializing it and applying it to something. So building to build is one thing, going to market on its own is one thing, but the really great founders are the ones who combine the two. It's also often we are told, oh, you know, Europe is distributed or Europe has smaller markets. Well, we're extremely smart and capable people who know how to take these perceived challenges and turn them into opportunities. So the fact that some of these smaller markets don't speak English means that they have to go global from day one and they have to look outside of their local market.

25:13And that creates a multinational global view from day one. So they understand how different markets operate. so that they can go to wherever it is that they want to go. They can understand the nuances of different markets and then find that simplicity so that they can go fast very quickly. I think these things are very important. It's also, like any other accelerator program, having access to other people who have done this before. So having mentors, having entrepreneurs who have built things that are coming and giving talks, having hackathons, organizing, having this builder mentality and not just this theoretical you're here to learn education but you're here to learn education to apply it to something you're not ordinary no nine to five meal deal lunches meetings about meetings compulsory team building living for the weekend you're made for more for the extraordinary and you need an extraordinary bank HSBC innovation banking exists exclusively to help visionary VC-backed tech and life science businesses achieve their ambitions.

26:19From pre-seed to IPO and beyond, we provide custom solutions co-created by actual industry experts to help you scale, delivered by a dedicated team you can call whenever you need to. Find out why four in 10 UK unicorns have chosen us at hsbcinnovationbanking.co.uk. If we really zoom out, what would be two requests you would make of European policymakers or heads of state or whoever to really enable Europe's AI community to compete on a global level? So first of all, I'm not a policy expert and I don't in policy. I work with founders and developers. So I think that's more of a question for one of my colleagues.

27:05But what I am already seeing, and I think the direction of travel is to just have the developer community across the countries be able to more seamlessly work together. As I was saying earlier, whether it is in hackathons or meetups, or even the universities that have students from multiple countries that end up in one place, the sharing of thoughts and point of views, I think, is only additive. When it comes to M &A, what does your crystal ball tell you when do you think we're going to see a sort of sweep of M &A in amongst Europe's AI startups and who do you think would be the buyers? I don't know about the when I can say what I would think would happen is that historically whether in SaaS or in the earlier days of AI we were seeing companies that were building great applications, but now for them to become true successes and like real businesses, you need to have platforms.

28:13So I tell founders now, if you're building something new, don't think about the point solution. Don't think about the unit economics or cutting a step in a workflow. Think about the owning the user and that user journey and experience as a whole. So I think some of the companies that are operating in very similar or adjacent spaces may end up coming together to build these platforms within a vertical or a workflow. And do you think, how much do you think big tech companies might be interested in buying startups or in what instances do you think it would make sense to acquire startups? I think it depends on the which big tech company and for what, but we're already seeing some M &As.

29:01We as a company really believe in the future of AI and science. So that's why we introduced Prism and we made an acquisition there in something that we thought was a great tool to help the users that we already have and we want to be supporting them. But beyond that, I think the market is going to define that a little bit more. you often take your daughter yes hackathons what does she make of it all so I've worked in tech forever now and I've been raising my daughter in tech and I think if you want your kids to learn about tech to learn about AI at this point in time I try not to teach her sort of from the beginning and first principles down as if she was studying a book or was in class but to start backwards from What projects are you excited about?

29:54What things would you want to build for yourself? And let's work backwards from that to figure out what tools can do it for you. And part of seeing the magic of what is possible is actually being exposed to other people who are doing it, which is why I say, again, go to the hackathons, go to the events, build in public, because A, you're seeing what's possible, but B, you get inspiration of, I can do this too. And C, you can maybe find other people, you can co-found, you can go on your own journey. For my daughter, what I was extremely proud about was that even from the very first ones that she came to, she would go to people, I would say to her, go and ask them what they're building.