S9E4: Guillermo Rauch, Co-Author of Next.js, is Revolutionizing Web Development with Vercel

I noticed that if I improved the internal tooling to ship changes for our startup, morale would go up, bug fixes would go out faster, we'd get more confidence, more psychological safety, more of everything that you love about your job.

Welcome back to Season 9 of The Room Podcast. Claudia, happy Tuesday.

Happy Tuesday, Madison. Thrilled to be back behind the mic with you. Since last season, you've had a few updates on your end. You're now at Harvard Business School. Congratulations.

Thank you so much, Claudia. It's been a dream of mine for some time to attend, and I'm thrilled to be back on the East Coast this fall. You'll have to keep me updated on what's going on in the Bay Area news and all things Priv, where you're still co-CEO and co-founder.

I absolutely will. Well, if you're new here, we're two best friends building our careers around tech and entrepreneurship. Since August 2020, we've interviewed over 80 guests, including the likes of Michelle Zatlin of Cloudflare, Vlad Magdalen of Webflow, Ciara, the Grammy Award-winning R&B vocalist, Spencer Raskoff of Zillow, and many more. Through these conversations, we open the door to the room where it happens.

Our guest companies share a combined valuation of over $47 billion and have raised north of $6.5 billion in venture funding. And last year, Spotify named us a top 5% most shared global podcast. This season, we're back with 8 new stories touching on everything from open source, trust management, generative AI, and more.

If you're based in San Francisco or Boston, catch us IRL this fall. The best way to keep up is to subscribe to both the podcast and our newsletter, which you can find at theroompodcast.com. Before we dive into today's story, we have a short message from our sponsors. Silicon Valley Bank, now a division of First Citizens, remains fully committed to serving founders, executives, and investors across the innovation ecosystem. As the largest banking team focused on technology, life sciences, healthcare, private equity, and venture capital, SVB is uniquely positioned to deliver specialized lending and financial solutions clients need to reach their ambitious goals. Find out more at svb.com.

Cooley is a global law firm built around startups and venture capital. Since forming the first institutional venture fund in Silicon Valley, Cooley has formed more venture capital funds than any other law firm in the world. The firm has 60+ years working with VCs. Helping form managed funds, make investments, and to handle the myriad of issues that arise through a fund's lifetime. Learn more about the firm at cooley.com and also at cooleygo.com, Cooley's award-winning free legal resource for startups.

Welcome back to another episode of The Room Podcast. Today, we are excited to share our conversation as we sit down with legend and innovator in the software development space, Guillermo Rauch, CEO of Vercel and creator of Next.js. Next.js is the most popular React framework that most of your favorite websites and apps are leveraging, and Vercel was born out of the development of Next.js. Vercel's frontend cloud gives developers frameworks for workflows and infrastructure to build a faster and more personalized web. If you have been to the websites of Adobe, eBay, Loom, Zapier, even the Washington Post, you have encountered Vercel. To date, Vercel has raised over $313 million in venture funding from the likes of Accel, GGV, CRV, and many more. In today's episode, not only do we walk through pivotal moments in the history of frontend development and the building of Vercel, but we also discuss key themes such as when to compete with a platform versus building as a layer on top. The art of making shipping software addictive and evaluating, then executing on what you're uniquely positioned to solve for. Let's open the door.

Guillermo, thank you so much for joining us today here in the room.

Thank you for having me.

We're really excited to share your story and dive into the future of Vercel as well. But before we go down that rabbit hole, we like to start at the beginning with all of our guests, and we're curious, Where did you grow up and how has that shaped your view of the world?

I grew up in Argentina, in the outskirts of Buenos Aires. Our family came from pretty humble beginnings, but both my parents were engineers, chemical and industrial engineers, and they shaped a lot of my worldview and inspirations for science and technology and doing great things, learning English. We got a computer in our home pretty early. We got internet for that computer. That led to how can we hack things around? How can we make systems evolve in different kinds of ways? And I got really into engineering and open source, and it became the inspiration for my life to build things on the web, to then enable other people to build things on the web. And it became a career. So I actually dropped out of high school. I got my first job out of Switzerland, so I kinda went from Buenos Aires to Switzerland, then to the Bay Area, and now I'm in San Francisco, and I started two companies, one that I sold to Automattic, and now I started Vercel and the Next.js framework.

So you, from the beginning, were always a tinkerer, and that seems really core to your family's identity, having parents as engineers, and realizing you were in this time and space where access could become unlocked through the internet, and you were an eager beaver to get ahold of that access for yourself and for others. In that vein of early identity though, at the time, did you always think you were going to become a founder yourself?

Looking back, I didn't think of it as a plan, but I remember my first side hustle was I would burn CD-ROMs with music and sell them at my elementary school. This is prior to me learning that I could engineer things and then sell my software and my services and things like that. And then after I started building websites, I actually created a network of websites to create refers for the eBay of Latin America, which is called MercadoLibre. So I always had this idea, turn it into a business idea, turn it into a business. But I think the thing that made it all really work and connect for me is I had a passion for, to your point, enabling others. And when I came to the Valley and I saw this emergent trend of like, you can build businesses on top of open source in dev tools, that was my true call. But it all starts with that passion for tinkering. I don't know, I just never wanted to have a boss and things like that. I think there's also some personality traits there as well.

I think every good founder has had that thought. Like, I'd like to be my own boss. This is where this is going. You mentioned Switzerland. I was going to jump into MooTools. How are they connected?

I started contributing to this library for UI development called MooTools. The appeal of it was with MooTools, you could make certain things possible in a web browser that the more traditional sort of backend-oriented Java applications could not do. Those apps could not do animation. They could not do real-time data synchronization, they could not do something at the time was called AJAX. So the enabling technology there was JavaScript, which runs on the client side of the web browser. And a lot of fields in engineering started with really smart people just write the code. Then they realized, oh my God, I'm writing the same code over and over again. So they're like, maybe we should put this into a library. And then a lot of the expectations today at these libraries are open source. So MooTools was one of those pioneering libraries for UI development. And I was very lucky that I got involved with it very early on. I became a core developer of this basically global group of enthusiasts. It turned out that those enthusiasts became really prolific and influential folks. Facebook decided to hire a lot of these engineers because they saw in MooTools a lot of promise for how they wanted to run Facebook.com's frontend. And then a company out of Switzerland also bet on this technology for their startup. And they were like, yeah, we need to hire the experts. It turns out most of the experts went to work at Facebook. There's this one expert in Argentina who's 17 years old. That was what the consultants basically sold them on. They took a bet on me and it turned out really well. I was able to contribute a lot. They opened an office in the Valley and that kind of led to me visiting SF and, and doing more things.

Oh my gosh. And that's just really early days of, I would call it like user-friendly light front-end experiences. It's not what we know of today, but it was certainly better than what it was in the '90s. Talk a little bit about that. What was it like moving both across the world to Switzerland and then back across the world to San Francisco in this kind of post-dot-com boom and bust era, but still so much promise in the internet.

Super interesting that you mentioned step functions because one of the inspirations that I had when working on Mood Tools was there's this new category of application that is very dynamic. So for example, Gmail had just come out, Facebook Chat, those floating little chat things on facebook.com had come out. Google was starting to autocomplete and suggest stuff as you typed, even triggering the search as you typed. A lot of things that to your point, we take for granted now, we're like, holy crap, this is impossible unless you're Google. So part of my effort was like, how can I share access to this incredible superpower to everybody? So that's post.com. Now I remember vividly when it first arrived in SF, we're technically in a crisis, right? Like it was a real estate financial crisis. And, but to me at the time, it was just so focused on the pursuit of, again, continuing to enable the democratization of these incredible technologies. A lot of what I feel happened at the time is very similar to what I'm feeling now about AI. Yeah, there's some indicators, there's a lot of chatter about the markets and things like that. But once again, we have this step function of what happened to like the social web and Web 2.0. Now we're seeing that with AI. History evolves and changes, but a lot of things still seem familiar and ring true to how it was back then.

So you're in San Francisco, you are on this pursuit of building better tech to enable more on the web. And as you alluded to, Vercel was not the first company you started. You founded Cloudup. An offshoot to edtech company LearnBoost, which you were also co-founder and CTO of, which helped people easily share files. As you mentioned, CloudApp was acquired by Automattic in 2013, the company behind WordPress, to power their editing and site-building technology. Tell us a little bit more about this early founding experience and also what advice you might have for founders who are thinking about acquisition opportunities.

That's super interesting. Number one is I think looking back, we started out with edtech, but what I really wanted to do was enable folks to share with no friction. And again, like I had such a strong passion for enabling developers. So looking back, I should have taken a shortcut to what eventually became Cloudant and maybe even eventually became Vercel, which is I was pursuing what I think I was uniquely positioned to solve in the world. Whereas the edtech thing doesn't make sense because looking back is like, I dropped out of high school. I'm the poster child for non-traditional education path, right? So honestly, it could have been a coding education thing of sorts, but it was still a really valuable experience because of the network that I developed and the folks that I got to meet. And being CTO at such a young age was also incredibly foundational. In terms of everything that I learned and the kinds of problems that I was exposed to. And even just the, one of the things that I look back really fondly of is I developed this obsession with CI/CD, so continuous integration and continuous delivery or continuous deployment. I noticed that if I improved the internal tooling to ship changes for our startup, morale would go up, bug fixes would go out faster, would get more confidence, more psychological safety, more of everything that you love about your job. Like I come in, I can be creative. I don't fear about deploying on Fridays and outages and things like that. So having great pipeline for software delivery. So even though we're in edtech, I was excited to like the idea of Vercel. I noticed every time I would be in a position to sort of lead an engineering team and set up the foundations, I would always focus on that. Like, what is the platform that enables our innovation to happen? Do we have the strong foundations in place? The next big realization was I love technology that reaches a huge number of people. And there was a great feature within Larmbo that enabled really fast file sharing, and it made it a collaborative thing to create links that contain content. So we simplified and we made the product more general. And then I had the opportunity to meet with the founder of WordPress. I showed them all this technology that we'd created. Simultaneously, they were also arriving to the conclusion that they needed to invest more in their front end, in JavaScript infrastructure. And I had always had an immense amount of respect for WordPress's roots in open source and the kind of business that they were building. So what I would suggest to founders that are confronted with this challenge/opportunity, because it's always hard to say, here's my thing, take it, or we're ending our roadmap and things like that, is I looked to the possibility of learning from some of the best people in the world, Matt had in many ways created the category of company that I would one day want to create. So the ability to be able to partner with those kind of people is incredibly alluring. And then ultimately the other thing is I wanted to be set up for success in a more direct way in the future. What I mean by this is I think edtech was probably not my bright focus. and I knew that the next time I would do it, I would do something that's very authentic and true to me from the start. And that's how it eventually led to Vercel.

Thank you for sharing that experience and how you thought about the acquisition to WordPress. But certainly you were still coding away after CloudApp concurrently to being a startup founder. You were also closely involved in creating numerous influential open source projects like Socket.IO and Mongoose. Really with the mission, as you mentioned, of creating tooling to enhance speed and agility for developers. Also psychological agility as well. We've had some bad deployments recently and it's very stressful. Tell us a little bit more about how Socket.IO and Mongoose have changed the development landscape today and what you learned through co-authoring and contributing to those projects.

I'm very proud of those projects because, first of all, there is a reach Even though now we've created something that's far bigger with Next.js, I still go to meetups or conferences and people tell me like, because of Socket.IO, we're able to create this startup or sell it. Or I always tell people the engine of real-time communications that I built for Socket.IO still powers Notion. So Notion's real-time syncing engine is the engine I built back in the day. What I learned really though, that's really transcendental, I think, is they focus on developer experience. So you can create the best technology in the world, but if it's not accessible, if it's not streamlined, if it's not easy to learn. I developed a lot of tricks. I remember with Socket.IO, I was obsessed with what I later would call like README-driven development or documentation-driven development where like I started by explaining what the tool does. What the basic shape of the API looks like. I outlined some of the benefits in a few bullet points, and that's my one-pager before I write even a single line of code. A lot of the projects that I've created have come backwards from the experience that I wanted to enable for engineers. And that was certainly true for Socket.IO. Mongoose similarly was a foundational project for the MongoDB world. It's still very widely used for a lot of projects in the JavaScript ecosystem that use Mongo as their database. And again, the thing that I was learning is like, look, like open source is positioning itself in the critical path of some of the biggest businesses in the world. I actually saw Mongo from, I met with the CTO actually as soon as I moved to SF, because he was like, oh, like this guy's created Mongo. Talk about open source opening doors. I actually built an incredible network in weeks because I remember I met Jeff Lawson from Twilio right away because he was interested in Socket.IO. I met with the CTO of Mongo because of Mongoose. So it was my content, my work, my APIs that were opening all these doors. But then I was also noticing how you can actually create transformational generational businesses that start with those communities of developers, that start with developer experience. Then I saw Stripe get created. So all of that was sort of inspiring me to, once I basically left WordPress, I said, how could I reproduce the ideal infrastructure on top of which you can create the best products on the web? And again, I turned to open source when I created Next.js and then the infrastructure to deploy it in the cloud with Vercel.

So let's dive a little bit deeper into those experiences. So soon after Mongoose, Socket.IO, Next.js was born. For our listeners who might not know, Next.js is the most popular React framework that most of your favorite websites and apps are leveraging. My startup, our entire app is based on Next.js. So big fan. Tell us more about the historic co-authoring of this framework. And how you drove adoption really to get it to be the most used web framework. Talk about impact.

When I discovered, you called Next.js the framework for React. When I saw React come out of Facebook open source, I recognized, it took me a few weeks, maybe a couple months even to recognize that it was going to be a different way of building things. I actually look at large language models the same way today. I recommend to people, sometimes you think, oh, I have to invent my own React. I have to discover relativity. I have to create my own ChatGPT. Sometimes what you really need to do is understand that certain technologies are just generational. They're poised to change the world. And the right thing to do is build on top of them and/or become an expert in them. So the first thing I tried to do when React came out is I wrote my own React. And I realized that there was not going to be a point in creating another one. I wanted to do that because I like to learn that way. I like to rebuild things. If I cannot build it, I cannot understand it. So I tried to build my own. I did build my own. And then I realized what the world actually needs is to be able to bring this technology to powered or actual websites and web applications. An engine is not really what you're after to solve your day-to-day. You want the entire car. So Next.js was my idea for like, how can I democratize the way that companies like Facebook and Google actually build for the web? They have internal frameworks that are extremely powerful. They're not open source. They do things that still feel alien to most people, like running experiments at scale and being able to say, this UI, we can't just risk all of google.com search traffic on this new idea that we have. What is the time from idea to experiment? Can you render that dynamically and efficiently at the edge so that it's fast and it doesn't render a bunch of spinners to the user? So I just wanted to build a framework that really enabled that kind of like world-class product development. And very quickly thereafter, I realized maybe almost even simultaneously that you can't just give people an open source tool that runs on their laptop today. You have to actually bring them to the cloud. You have to give them the platform on top of which they can collaborate, they can ship, they can get insights in real time. They can remove a lot of the undifferentiated work of running a website at scale. Because when you join Google or Facebook, they don't just give you a pile of code. They give you incredible platforms that enable you to ship at scale. And again, they put you into the pit of success. They don't give you SSH so that you log in into a server and you delete Google by mistake, right? Oh, I deleted the database. So we set out to really democratize access to all these technologies and Vercel became the platform on top of which, to your point, a lot of startups, a lot of innovations are being created and deployed on the web today.

Vercel, as you mentioned, really gives developers the ability to build more. I'm a Vercel customer, so incredibly excited to be chatting to the founder, but tell us a little bit more about the aha moment that led to creating a company that is Vercel and really taking that product to market.

One of the aha moments was I realized that most companies become really good at deploying the main thing that they become famous for. If you're a large hotel chain and you've been online for 20, 30 years, you've become really good at deploying the one thing that you have today. What's really hard is to keep up with the relentless change that happens on the internet. The best practices, the platforms, the tools are constantly changing. I realized that if I could somehow outsource what we'd call the developer infrastructure of your company and make it a really delightful and also very flexible experience, because on the other side of this is I've been burnt many times as a CTO myself where something promises me simplicity and then it's just not flexible enough. So I wanted to craft a platform that was durable, that really large and ambitious businesses could build on top of. The classic story that we tell is that when companies ship on Vercel, they don't go down through Black Friday. You would be surprised about like how many incredible brands still struggle to scale on the web. So Vercel is really enabling that transformation to happen. We in the industry talk a lot about digital transformation, but now being on the other side of supporting these large brands, I see the pains that they go through. A lot of folks that have been born in Silicon Valley, figuratively, like we've been spending so much time here. Sometimes it's hard to develop the empathy for what happens to the brands that have been around for 100 years that have seen their businesses be transformed overnight. Because what do I have to do next? Do I have to create a ChatGPT plugin? If I'm Coca-Cola to sell you my product, how do I gain that agility in this ever-changing world? Vercel is that idea, really. Of course it's a technology. Of course there's Next.js and we streamline its deployment, but it's really getting at the core of what makes a business succeed online that we enable.

I love that you're drawing out this tension between legacy brands or companies that have been in the ecosystem for perhaps 150 years and their journey of adoption for these technologies. Having been a product manager at Gap under our migration from on-prem to Azure, I saw firsthand what that workload was doing both to our engineering base as well as the contractors that we had helping make it possible. And it's something I don't think the average entrepreneur has that empathy that you're alluding to of, oh, you can just implement our plugin. We have these API keys, like it's no problem.

I love that callout. We were getting into what advice would I have to entrepreneurs today is you have to simultaneously be a lighthouse. Fundamentally, companies are coming to us because we're creating this foundational technologies. They want to work with the experts of the web, but simultaneously you have to have the empathy to understand where are they at today? How can they facilitate an incremental adoption of our technology? I think a lot of folks get caught up in like, look, I built the ideal. But the deal, when it meets reality, it might just have this impossible to understand path to getting into production. One of the things that I became very obsessed with early on was this idea of how can we position ourselves in the critical path? I don't want to be this sideshow utility that makes your business 5% better. We need to make your company 10 times better in order to succeed. And that becomes a race of how do you get into the critical path really quickly? And to get into the critical path really quickly, you can't just come to market with like, welcome to Vercel, replatform everything on us in order to get a taste of our success. And so a lot of that was developing that empathy with the kind of brands and businesses that you're talking about.

And it's funny too, just like to add on there, is the brand that has been around for 50 years has been burned before. Bore by other platforms telling them that they're the future of the web, and you have to go against those experiences and past failed deployments to get to the truth of what we know to be the best platform for these brands to use.

Yeah, and by the way, this is a very real thing, right? Like, everyone is always out there just promising you the world. Like, here's 3 easy steps to get 6-pack abs, and here is this streamlined diet, is the caveman diet 3.0, and like, Here is a platform that's going to solve every problem. What we try to root ourselves in, and this is again going back to my DNA as a frontend engineer, is number one, we position ourselves very incrementally. We're the frontend cloud. You can add Vercel to whatever you already have. Number two, we prove our value with measurable returns on investment. The frontend to us is a profit center. You make it faster, you increase conversion. You make the DX better, you decrease your bottom line. You just make your entire team happier, more efficient, they ship more code. So we try to root ourselves in like, what are the measurable, realistic advantages that we bring to the market?

I really want to talk about go-to-market, which we're touching on briefly, but before we do, just for our listeners who might not be as familiar with this concept of an open source platform like Next.js, then launching a business behind it with a hosted solution. Could you just walk us through a little bit more of the meat and potatoes behind that process going from Next.js to Vercel?

So number one, and the reason that I love open source is that A, it's becoming the buying preference for engineers, right? Like I use PyTorch, I use Next, I use now the AI SDK by Vercel to build my AI apps. It's just a viral way of establishing product market fit really quickly if you have what it takes, because you enter this awesome marketplace with zero friction for adoption. So I think it aligns the incentives between company and audience extremely well, not just on the, is the product any good, but also on like, look, Next offers you everything out of the box. And we don't constrain any features. That means we have to be so good to earn your business at streamlining its deployment, its workflows, and collaboration. Another way that we sometimes talk about Vercel is like, it's like AWS for the frontend. We have to compete head-on with every other alternative that exists for building a website today, which frankly is like a gazillion. So I think the transition was very clean because Customers today want services, not downloads. Downloads is how I was buying software in Argentina in like 1998. Like you would go to a website and the call to action was download our stuff, get a license, and then it's your problem. You run it. Not my problem. I can sell you some support, but you run it. Now I run it and I take the burden of patching, upgrading its security, ensuring that it keeps up with traffic and attacks and gets continuously improved. It's low latency. It has a global footprint. It's streamlined its connectivity to your data sources. It has all the integrations that you need. So that's the burden that we take on behalf of the customer. So the engineer gets the developer experience that they want. They want to adopt software that's open source. They don't want to lock themselves in. The other beautiful thing about Next.js and why it resonates so much with enterprise buyers is Yeah, there is other ways of adopting serverless technology. There is other things that you could be doing, but you're either writing your own frameworks, which is definitionally a pretty bad job to be done for the average enterprise. Like, oh yeah, stop. I'm going to write my own framework before we start this project. And the other one is frankly that, again, you want the optionality. If I have Next, I can take my workload wherever I want. World. So a large inspiration there was what Kubernetes kind of did for backend and infrastructure, but could we do that for the frontend world? And as you spoke earlier, in terms of downloads and reach has been, it's been incredible. Everything from like ChatGPT to TikTok to your favorite e-commerce website, they probably already run on Next.js.

Let's talk about some of those customers. So including Pryve, Claudia's incredible company, to larger enterprise clients like Adobe, eBay, Zapier, even Okta are all using Vercel to work and power their frontend. How did you go about unlocking those first enterprise customers and what advice would you have for founders who are also looking to make that leap?

There's a lot of controversy around like product-led growth and is that the right thing for me or for you, whatever. I'll tell you my take on this is that if you're in a sufficiently large market like we are, product-led growth has been incredible. We give the product, we put it into the largest number of people that we can. We make it extremely easy to adopt. Basically, the adoption pattern of Vercel is you use Next, Svelte, any frontend framework. You're already pushing that code somewhere, like GitHub. and we basically ask you to connect that repo to Vercel, and your project is live. There was no platform engineering work. You didn't set up a CI/CD pipeline. You didn't set up any hosting infrastructure. It's just completely streamlined. So product-led growth has worked really well for us because projects are constantly being started on the web, whether it's enterprises that are starting new projects Or whether it's that engineers are always trying out new things. And that's why we have over 3 million deployments per week and 40 million downloads across all our open source projects every week. To give you an idea of the adoption pattern, you link that repo, you push code, we deploy. So when I talk about that inspiration from LearnBoost, over the last 24 months, The engineers on the Vercel platform have shipped 153 million times because we just make it, basically we make it addictive to ship software and that creates a great foundation for product-led growth. Consequently, what ends up happening is that most CTOs, most VPs of engineering, most principal engineers, most senior engineers, most junior engineers at every company you can name are using Vercel or Next.js in some capacity because we've made it so easy for the product to be adopted. Now, some of those might be qualified to be enterprise customers where Vercel kind of gives you like the premium top-notch service, extra support, SLAs, more security capabilities, and many of those stay on the self-serve path of Vercel forever. So we run this two-pronged approach to how we partner with companies to enable their success on the web.

I'm just really excited about how you are unlocking the speed of deployment in an era where the consumer expects speed. So to have a platform and an experience empowering faster consumer experiences through the underpinning of the philosophy that you should be deploying faster really makes a ton of sense and clearly is empowering hundreds, if not thousands of websites every day. Congratulations on getting to the milestone of the business itself. And alongside that, when launching Vercel, you went down the path of raising venture capital to build and grow this solution. You've raised over $300 million at a $2.5 billion valuation with incredible investors from GGV Capital, Accel, and Tiger Global. Truly incredible. I've watched the tenacity it's taken Claudia and can only imagine you to raise this capital to build your business. We're curious, as you thought about getting this off the ground, who was the first person to say yes to investing in Vercel?

Funny enough, the first person was not only just a person that said yes, but a person that was rooting for me to start my own business. So Rafael Corrales, who was an investor at CRV at the time, now he has his own fund, was that person that first of all, had an eye for what open source is going to do for the world way ahead of the competition, I would say, because now it's very common to hear analysts, investors, all kinds of people say, let's look at the GitHub stars and there could be a business there and things like that. Rafael was a huge believer in that as a distribution channel, as a way of going from zero to one for new internet businesses. So That was quite awesome to have as a partner in the early days. And then CRV and Excel were also the early foundational partners of the company. So Dan Levine at Excel and Reid Krishnan at CRV, again, huge believers in this. Dan has invested in other open source businesses like Sentry and EdgeDB, and Reid has done the same. And Reid has a great tagline at CRV that he calls like power to the developer. Which is also like a huge underrated trend is that I was lucky enough to be an early investor in Auth0, which got acquired by Okta for $6.5 billion. And one of their early taglines was also the developer's new CIO. If you make this technology so easily accessible, of course we talk to CIOs all the time, but there's just a huge number of really leveraged individuals within the organization rooting for the adoption of your technology.

As you mentioned, with GitHub Stars and telling your story through open source versus enterprise, and I'm a founder and I'm either sitting on the open source side trying to go enterprise or vice versa, how would you think about fundraising? What, what's your advice to those founders on the precipice of that decision?

The important thing is that it's not just something is open source for the sake of it. I think it has to make sense. I called out that Next doesn't have any limitations. It's not a funnel into our business. In many ways, it's something that enables us to do our best work. We dogfood Vercel extensively to build vercel.com. So there's all this awesome fitness functions that we've set up for ourselves as part of— Next is an element in that world, and it certainly helps us. It helps the community. So it's this win. In terms of should you start a company like this, I think one thing to note that is real talk is building infrastructure for the internet is really challenging. It's the most crucial part of anything that a company purchases today. If your site is not online or if it's slow or if it's whatever, you don't exist. So I love that we positioned ourselves in the critical path foundationally. But of course that takes a lot of time. You mentioned the rounds of financing. It's hard stuff. You need a lot of capital and you need a lot of time to harden a platform like this. It's not a warning for like, oh, this is not for the faint of heart, but I think it's also just being realistic about what it takes today to run planet-scale infrastructure for the internet.

The perfect dovetail into my next question. As you mentioned, it's not for the faint of heart. Being a founder is full of high highs and conversely low lows. Could you share about a moment in your journey building Vercel when things just did not go as planned?

Going from zero to one, extremely difficult. The number one thing was recognizing that sometimes you have to say no to certain customers that you're just not ready for. Regardless of how much excitement exists on both sides. So I think the tendency for an entrepreneur would be like, hell yes to everything. I would say actually saying no to things that are compelling and alluring is more important in many ways than saying yes. You're going to remember a lot more times the things that you turned down and you decided to focus. So when I look at things that could have been less painful is How can we get to focus faster? Having a project that goes viral like this just opens up every temptation in the world to do absolutely everything in their discernment. So I look back and say, let's get to ground reality faster. And this is my number one piece of advice to anyone that I advise, invest in, help out. How can we get to ground reality faster? How can you front real-world feedback about your technology faster? How can you iterate on that feedback faster?

Getting out of just the stories that we tell ourselves and our investors and really get to the truth of, is someone going to use this product and does it solve the problem?

Yeah, I think of it as retroactive pain. It's like the instant feedback of like, oh, interesting. I should not touch that hot stove.

Absolutely. Well, thank you for sharing that piece of advice to our listeners. You've mentioned AI a few times. Not only have you been along for stepwise functions in technological development, but we are all in the midst of one today. With the rate of development and breakthroughs in AI, companies and teams building, shipping, and deploying faster than ever, I'd love to hear your take on the future of AI and development and what's next on Vercel's roadmap.

A few years ago, I became interested in the definition of AI. This was before GPTs and large language models. I was like, what the hell is AI? And I remember reading this awesome article by one of the people that people think could be Satoshi. Like, it's the conversation. It's like, oh, did that person invent Bitcoin? Nick Szabo, who's defining AI as basically not a moment in time in which our intelligence gets replaced by a singular machine, but AI is basically this never-ending roadmap where every human skill one by one gets deleted and transferred and outsourced to a machine. So the best example would be just a matter of time before we don't drive anymore. I actually joked, I'm from Argentina, never learned how to drive because every car was manual. And I was like, That's over for me. So I came to the US, God bless America, I was able to drive very quickly because I had an automatic car. And I remember joking at the time, like, should I even go to the DMV? I'm on the cusp of not having to drive at all. Please, can you accelerate? Because I don't like going to like any bureaucratic process. So especially the DMV, although it has become better recently, but really I don't want to drive. So like, that's an example of a human skill that is like just getting deleted. And it's not going to be a singular piece of software that runs everything that deleted that. But now look at the job of reviewing copy and finding typos or redundancies or improving the grammatical expressions or finding repeated words. Like now that's outsourced to ChatGPT or something like it. So one by one, everything that is fundamentally not creative that we could automate away, we're going to call it AI. I actually think realistically it's going to be a mix of AI systems and other kinds of heuristics and software and even human intervention. It's going to be complex hybrid systems. But what I'm excited about is that now it's being recognized that our job and the fundamental innovation engine of Silicon Valley and startups should be AI. Because I think it just like laser focuses us on— it's the right vector of innovation, which is to again, automate everything to make our lives better. With that kind of disclaimer out of the way of I don't believe in one true AGI that is going to solve everything, etc., I do think it's a series of exciting new technologies and foundations above which you can build the next series of innovations. And this is both true for startups and enterprises. One of the things that I'm particularly excited about now is, okay, a lot of amazing foundations of AI have been created. As I talked about earlier, when React got open source, the other React moment that I'm seeing now is these APIs that give you this instant reasoning machine out of the box, a couple API calls away. So that's the OpenAI API, Anthropic put out Claude 2.0 today, there's Hugging Face, there's all the open source models, there's replicate.com, all of this amazing technology now exists. It's now time for that other side that I talked about, that other layer of software and workflows and complexity and humans in the loop and all those kinds of things to be built in the service of creating amazing products. and that's what Vercel is focused on enabling. So we just open-sourced our Vercel AI SDK. It makes it really easy to connect to these reasoning engines like OpenAI. We gave you an example. It's called chat.vercel.ai. It's basically your own ChatGPT that you can clone and deploy, and this could serve as a foundation of products. It could serve as a foundation of ChatGPT for the enterprise and like having your own version behind login and extended and remixed of that kind of technology. And again, we think that the future of the web is just creating better, faster, more delightful experiences, not the undifferentiated infrastructure. So our go-to-market here is pretty simple is there's maybe 3 or 4 companies in the world that are going to train this large language models. Estimates range from like, it costs like $50 million to $200 million or whatever to train them. Most startups and enterprises are not going to do that. They're not going to build their own cloud infrastructure, but they are going to have to include it into their products. I'm completely convinced about that. I was convinced in 2018 when Copilot came out and it gave me this ghost text, this suggestion when you code. I was like, the whole world will need this. And I'm not just talking about writing code. We have to suggest what's next for every task that we do because it solves like 80% of the problem, right? Maybe not 100%, but it helps you. I think now is the time of the best time, probably in a long time, to build those kinds of products sitting on top of those foundational technologies.

We're so excited for just the future of what the AI application layer is going to look like. Thank you for sharing kind of your POV on the space. As we come up on time, I'm curious, we chatted a lot about what's next for not only Vercel, but just AI and the development ecosystem more broadly, but what are you excited for this year personally, Guillermo?

For this year, continuing to deliver on our dream of what we call dynamic at the speed of static. So what we're seeing now, in broad strokes, and AI is actually a participant in the benefits of this infrastructure. Is the world is moving away from static experiences into dynamic personalized ones. And AI personalization is just one component. Democratizing the dream of every page is tailored for every user and making the web faster for every business on the planet. Like that's at the heart of what we're doing at Vercel. And there's a lot of exciting releases and innovations that we're bringing to market to continue to make that vision a reality.

We're excited for that as well. And we'd love to now ask our hero question of the podcast, which we ask every guest as we sign off from our conversation. Who is a woman in your life that has had a profound impact on you and your career?

I think a lot about the Ada Lovelaces of the world, the foundational women engineers that were in many ways behind the scenes of this revolution that definitely have not gotten the right credit for. I believe there was a book a couple of years ago that came out. It's actually women who are behind the computer revolution. Here's all the data. Here's all the facts. My inspirations are always great technologies, great scientists, and I'm always in the pursuit of their recognition. Lisa Meitner also comes to mind, who was Perhaps the woman behind nuclear fission, Marie Curie, after whom I named my daughter. So there's a huge list, luckily for us, and I think it's overdue that we recognize them.

Thank you for sharing that, and thank you so much for taking the time to share your story and what you're excited for next on the podcast. We really appreciate having you on today.

Thank you all so much.

Thank you so much for joining us at The Room Podcast. If you want more from The Room every week, subscribe to our newsletter at theroompodcast.com/newsletter. We'll be back next week with a new episode, Tuesday, 10:00 AM Eastern, 7:00 AM Pacific. See you in the room.

Silicon Valley Bank, now a division of First Citizens, remains fully committed to serving founders, executives, and investors across the innovation ecosystem. As the largest banking team focused on technology, life sciences, healthcare, private equity, and venture capital, SVB is uniquely positioned to deliver specialized lending and financial solutions clients need to reach their ambitious goals. Find out more at svb.com.

Cooley is a global law firm built around startups and venture capital. The firm has been devoted to entrepreneurs and investors, partnering with both to transform breakthrough ideas into successful companies. Cooley works with thousands of entrepreneurs and newly formed companies to ensure that they are structured for growth, and long-term success. Since 2005, Cooley has been ranked the number one most active law firm representing VC-backed companies going public.