Vercel CEO: Deployment should be instantaneous

Vercel CEO and founder Guillermo Rauch discusses his company’s push to bring streamlined workflows, continuous deployment, optimized frameworks, and infinite scale to every developer.

Vercel CEO: Deployment should be instantaneous

Guillermo Rauch is the CEO and founder of Vercel, a cloud infrastructure company that offers advanced deployment capabilities for front-end JavaScript, edge, and serverless functions. Vercel is a kind of new layer on top of cloud platforms that automates a lot of the application deployment work that used to be done by hand (most of which fell under the “devops” heading).

Rauch is a longtime contributor to open source JavaScript, having also created MooTools and Mongoose. I got a chance to talk to him about some of the ground Vercel is breaking, including simplifying infrastructure and streamlining workflows for developers. Rauch also shared his insights into running a successful startup and how to approach funding.

Matthew Tyson: It’s exciting to talk with you. I’ve been aware of your work for a long time, since MooTools. How did you get your start in coding?

Guillermo Rauch: I got my start in coding when I was 10 years old growing up on the outskirts of Buenos Aires, Argentina. We got our first computer at home, which opened the doors to a global web and has shaped much of my life’s work, including what we are building at Vercel.

I went through many false starts: Visual Basic, DJGPP for C/C++ on Windows, Borland compilers. Once I switched to Linux, coding really started working out for me. Open source opened the doors for me to a more developer- and beginner-friendly environment.

Most importantly, I was able to find community. I made friends in local Linux development groups as well as online forums and IRC rooms. Those folks, many of them anonymous, had a tremendous impact in my journey.

Tyson: Ah yes, I remember Borland compilers! How did you migrate from Linux user groups to the helm of Vercel? Was that a conscious decision? “I’m going to run startups!”?

Rauch: When I was learning to code, open source had not yet gained the momentum it has seen over the past decade.

Today, infrastructure is global and developers are not looking to operate the infrastructure, they’re looking to be amplified by it. The evolution of open source has also shifted toward the frameworks that power the application layer, rather than the infrastructure layer. Linux has become an implementation detail of what makes our infrastructure possible. The developer mindset has shifted into, “I want ease of use.”

For startups operating at a fast pace, this is how they can scale software and reach more people. Open source is fundamentally better for the web as the world is adopting and creating new technologies faster than ever before.

Tyson: Yeah, the pace of change is astonishing. 

Vercel is described as “The workflow for the modern web.” When should a developer think about looking at Vercel? In what situations should they begin thinking, “OK, the team could use Vercel to address this need.”?

Rauch: We’re seeing the beginning of a new era where front-end developers can self-serve to global, virtually infinitely scaling cloud infrastructure with zero configuration and no devops work.

The first point of contact with Vercel starts when you want to share the progress of your Next.js (or Svelte/SvelteKit or Vue/Nuxt etc.) application with the world. You can start deploying your app to the cloud and share/access it instantly with anyone in the world as soon as you start developing it.

Not only are we seeing infrastructure getting simplified, but folks are “syncing” their changes to live applications constantly. Much like Dropbox and Figma did for productivity and design, we’re doing for development.

Tyson: The idea of “live syncing” app changes is a big step towards the ideal and promise of continuous delivery. I imagine teams see a thoroughgoing improvement to application life cycles as they benefit from things like finding bugs closer to their source.

Rauch: We’re working to democratize the best practices that have been discovered and developed by the giants of the web. Companies like Google and Meta have mastered the ability to streamline workflows, where builds are instantaneous, the frameworks have been optimized, they are maintained by dedicated framework teams, and the developers work in large monorepos with billions of lines of code.

We’re bringing the power of those streamlined workflows to any developer, team, or company with Vercel.

Tyson: What are the areas where the Vercel team is pushing the envelope? The big areas of R&D right now?

Rauch: We’re pushing the envelope in three key areas this year. First, everything we’re doing around the edge—primarily in collaboration with partners—is focused on trying to bring data to the edge in making this instantly global approach to development.

Second, with Turborepo, we’re giving customers the tooling that scales to billions of lines of code from front-end code bases and democratizing the concept of monorepos to all of our customers.

And third is the complete elimination of the provisioning of a development environment, making development instantaneous, real-time, and browser-focused. Next.js Live is a key component of that push.

Tyson: Is keeping up with the astounding pace of front-end tech evolution a huge challenge?

Rauch: We built Vercel to solve slow, disjointed development workflows and instead integrate the entire development and deployment workflow into a single process. Our goal is to drastically reduce the time it takes for our customers to build innovative products and launch new ideas.

That’s why our focus is on staying ahead of the challenges that might arise for front-end developers and ensuring we are constantly building the infrastructure they need to build the best web experiences for their customers.

By providing developers with the SDK for the web, we are constantly curating the pace of innovation around the front-end ecosystem. We do this while also staying flexible and welcoming all the new technologies that are coming to market. It is our job to simplify this pace of evolution and package it to our customers in a consumable way.

Tyson: Do you think these kinds of capabilities will extend to other stacks? Will we see Vercel handling Java/Spring apps in the future?

Rauch: We’re focused on allowing the more traditional stack to modernize while speeding up the front-end layer. Vercel amplifies the developer experience at the edge—allowing them to move faster and make a faster web app—all while bringing data from back-end services. This is exactly what allows us to move so fast in the enterprise. Customers don’t need to port back ends to Vercel. The most we’ve asked them to do is decouple the monolith.

What we’re finding is that most companies are already on that trajectory. So we come in and we say, “Oh, you’re already doing that. Cool! Oh, you have a GraphQL API. Cool! Now connect to Next.js with the GraphQL API.”

Tyson: I find it fascinating to discover what makes tech startups succeed. What is your advice for others who want to meet with startup success?

Rauch: Vercel was very much born out of open source, building community from day one, and iterating very closely in public with that community. This is where most startups are headed.

They’re launching earlier and they’re focusing on building relationships with their early users and customers. They’re building together.

We have always been very focused on the customer experience. We built backward from the ideal experience that we knew developers wanted.

From there we got into, “How can we make this scale?” “How can we implement this?” “What things can we write ourselves versus piggyback on others?” But the focus was always the same.

Tyson: That is very interesting, the idea of starting with the experience you want to deliver and working backward into the tech. A lot of us coder types get that one upside down.

Another thing Vercel is doing is supporting both open source projects and its commercial enterprise. Has nurturing that synergy been part of the strategy from the get-go?

Rauch: Early on, we bet on the fact that open source would become the way that both the developer and the CIO prefer to adopt software today. Enterprises are choosing open source, and developers will not adopt and learn a completely proprietary front-end stack.

However, it’s not just open source. It’s also building open platforms. Our platform is extensible to tools, hooks, and integrations. We care about that as much as we care about open source.

Tyson: What advice can you share about the fundraising part of running a startup, for people who are looking to jump into that fray?

Rauch: I am a strong believer in building the foundation for a developer community and paying attention to their engagement with the product. Some investors look at how many stars a project has on GitHub as the primary way of understanding whether there is funding potential for the technology. This is a signal of how well open source and enterprise SaaS are resonating.

My advice for fundraising: Start by finding product market fit through open source. Before you get real currencies, see if you’re getting that other currency in the form of GitHub forks, stars, and npm downloads.

Next, finding the balance between open source and proprietary software is the tricky part. Rather than cutting features out of the open source package to sell them, find ways to amplify that project through the business instead, ideally through an autonomous service.

Tyson: Fantastic. Can you talk a bit about Vercel’s tech stack? What languages and frameworks do you run on, and which cloud platforms? What are some areas where these choices worked well or were challenges?

Rauch: As a general rule of thumb, we like to focus on open source and on an open source tech stack. Betting on open source has worked extremely well and has given us the versatility to choose the right underlying platform or solution for our customers.

Next, we’re big believers in multicloud and optionality, which is why we have big investments and close partnerships with AWS, Azure, and Cloudflare. The value we bring to the market is that we can choose the right underlying technology, on behalf of the customer, to maximize their security and minimize their latency.

In terms of languages and frameworks, we are big fans of using our own products. We use Next.js, TypeScript, Node.js, and Turborepo. We also use Go and Rust because the future of JavaScript and TypeScript tooling is native and because they are great system infrastructure languages for scale. This process allows us to validate our technologies before we put them in the hands of developers.

Tyson: There is a lot of activity (in JS and elsewhere) for improving build tooling. Vercel seems to be active on that front with its acquisition of Turborepo. Do you mind talking about that move and how it fits in with the vision?

Rauch: Front-end technologies are constantly evolving. The diversity of frameworks, whether it’s Next.js or Svelte; the diversity of languages, whether it’s JavaScript, TypeScript, or Rust; and the diversity of runtime, whether it’s JavaScript or WebAssembly. It’s ever-evolving.

Turborepo is a high-performance build system that abstracts the complex configuration needed for most monorepos into a single, cohesive build system, giving large enterprises a world class development experience without the maintenance cost.

This technique was pioneered by large companies like Google and Meta, but Turborepo makes it available to the masses and solves the problem with more versatility, getting customers as close to zero configuration as possible.

Turborepo allows customers to structure their investment into their front-end code base in a way that the underlying applications can pick and choose how much they share of the broader code base. Customers can try out new technologies, while preserving a system that scales with the evolution of their team and their code base.

So as code bases get really big, and as the number of different technologies that are involved in the repository grows, customers can orchestrate that complexity with Turborepo.

Tyson: These are great insights. Thanks for taking the time, Guillermo. One last, sneaky question: Are those of us in the public going to get a chance to buy Vercel stock soon?

Rauch: Building an open platform for everyone in the world is an ambitious goal. Right now, we are focused on the fundamentals: Building a product developers love and empowering anyone to participate in creating the future of the web.

Copyright © 2022 IDG Communications, Inc.