CircleCI NewsMar 5, 20257 min read

Why CircleCI? Or: Falling in love (with change) again

Rob Zuber

Chief Technology Officer

CircleCI logo on a green background with some graphic elements.

Change sparks innovation.

Changes = updates = value delivered.

Change is risky.

Change is exciting.

The wrong change can bring down your application… or possibly your career.

The only constant is change

There are many kinds of change that punctuate the life of software builders. There’s change you don’t want, like an unexpected library update that causes your pipelines to throw errors. There’s the drumbeat of constant change, like the annual impulse to recreate your JavaScript frontend because you think (you know) it can be better. And then there’s the promise of cutting edge change, adopting new tools like Cursor or Augment that offer faster and smarter ways of working.

However, if the cost of every change, every time, is a massive investment, you’ll be in trouble. Which is why I believe CI/CD should absorb this cost as much as possible. And my team and I built CircleCI on this premise – to lower the cost of change so you can do the work you’re here to do: making things better.

I know that you, our customers, want to be better, because we’re software builders too, and we want to be better. Change is the one constant in this work, and it will keep happening: from package updates to replatformings, and everything in between. While we can’t stop change (and frankly wouldn’t want to), we do want to make it easy for you to experiment – and that philosophy is behind every product decision at CircleCI.

Despite my passion for this, this isn’t any kind of new revelation. In fact, at every major technology turning point since our founding in 2011, we’ve helped you make more flexible, pragmatic, and strategic decisions about how you deliver software. Let’s revisit some of these moments together.

Docker

Docker was released to the public in 2013. But by 2012, we had already noticed our customers’ challenges with version compatibility, so we actually built our own containerization. It wasn’t easy, and it certainly wasn’t the most elegant solution. But – at that point, off-the-shelf containerization didn’t exist, and we knew it would help make deploying software better for you. So we made it happen.

(And since Docker’s emergence, we’ve continued to invest in making it ridiculously smooth and easy to run deployments out of the box on CircleCI in the Docker image of your creation.)

No point in sugarcoating it: the move to containerization was a pain in the ass for us, but a huge boon to you… and that’s exactly the way we like it.

Kubernetes

In 2014, orchestration engines were just emerging on the scene. Kubernetes was one among many: Apache Mesos, Docker Swarm, Amazon ECS, and more. Operations teams planted their flags, making their bets on which tool would grow to own the market. With 20/20 hindsight, we can see that Kubernetes was the winner and took all. But in 2014, 80-90% of teams picked “wrong”… and spent significant time subsequently migrating their orchestration infrastructure. We were there to help them.

No one knows what will happen in the next 5 years (and anyone who pretends they do should be looked at with suspicion). But regardless of what new tools and methods take hold, we will be there. We built CircleCI to offer you resilience against the operational drag of change (when you pick wrong, or even when you pick right). We want you to be positioned to capture the promise of the next emergent technology… with minimal pain.

Artificial Intelligence

ChatGPT came on the scene in 2022 and has since been joined by a swarm of other GenAI tools which, together, have upended the way software is created. Teams jumped on the promise of the technology for creating code, and the way we as an industry test and evaluate software-written-by-other-software is still catching up. That said, we are thrilled to see GenAI bring new opportunities to advance the field, create new efficiencies, and enhance innovation.

And, just as with the emergence of Docker, Kubernetes, and so many other new tools, we’ve been at work finding ways to make it easier for you to experiment with AI without compromising your confidence in the software you’re shipping to your customers. We are building a future where validating AI-powered applications is as easy as triggering a test suite off a PR.

The next frontier: managing and validating change beyond the codebase

Did the shifting winds of Kubernetes, Docker, VMs, microservices, and more cause me some late nights? Did it make me lose all of my hair? Yes, and yes. But I’m as happy about change now as I was then (hair notwithstanding). Change is exciting! And keeping change feeling exciting, not onerous and terrifying, is exactly why CircleCI exists.

My team and I come to work every day to solve the problem of change, so you can innovate, build cool things, leverage incredible rocket fuel like GenAI, and deliver massive value to your customers. Our work in building CI/CD that makes change easier is never over, and that’s a good thing.

Many teams, though, are not seeing the potential of CI/CD as a companion to bleeding edge software innovation. I talk to teams every day who have CI/CD pigeon-holed to a narrow, and outdated, use case.

Retesting the codebase whenever engineers make changes worked well when apps were primarily made from internally developed code. But we’ve known for years that the vast (and growing) majority of changes to an application come from outside the codebase. So why search for the issue in the diffs in the VCS, when we know full well that most changes are coming from beyond the repository?

I want to invite you to take a step beyond the code-centric approach to CI/CD. We architect our entire CI/CD system to validate changes across all parts of the application, regardless of whether they originate inside or outside the repo. If it can affect your application in production (libraries, OSS, 3rd-party tools, LLM updates, and more), it ought to be able to trigger a pipeline. And on CircleCI it can.

I talk to software teams every day, and the teams I see who are succeeding at riding the waves of change, staying agile, and shipping steady value in unsteady times have a fundamentally different orientation to building software. They don’t just check the box on CI/CD; they are pragmatic and want real validation signals, fast.

We’ve always seen CI/CD as a kind of “cheat code” for building, learning, and improving your software delivery lifecycle, in a virtuous cycle that improves upon itself over time. We’ve always believed that good, strategic implementation of CI/CD can mean the difference between gaining a competitive edge in an uncertain market, and being lapped by competition that’s shipping updates orders of magnitude faster than you are.

It’s that important.

If you’re not using CI/CD as your competitive advantage, you’ll fall behind

Some teams may be stuck in a narrow CI/CD mentality because it feels more convenient or familiar, or because the pressure to consolidate tools onto all-in-one best-at-none solutions. But this approach leaves them scrambling to troubleshoot at 2 am when the next change causes their app to break –- even if their code works perfectly. This leaves teams bracing against the wind of change – and this stance not only feels bad, but goes against our grain as engineers. Teams who make this choice, implicitly or explicitly, are incurring tradeoffs they can’t afford, and will continue to be bogged down by reactive firefighting, poor user experiences, and slow feedback cycles.

For those who are always changing, always trying to be better, and creating better systems to foster innovation, we built CircleCI for you. CircleCI enables teams to innovate confidently by validating change across the entire application ecosystem, including external components and dependencies. Forward-thinking teams recognize that when a large part of an application’s value is delivered outside of code, they need a CI/CD orchestration layer that is also optimized beyond code. Software itself only has value if it is successful in production. By adopting CircleCI, teams unlock the full, Technicolor power of what CI/CD can really do, and gain the freedom, creativity, and confidence to broaden the scope of what they can safely incorporate into applications to deliver value.

Copy to clipboard