Opening Remarks — Day 1

Host Intro (Gene Kim)

Good morning. Welcome to the second Enterprise AI Summit here in Charlotte, North Carolina. I'm so grateful that you're here, and we have two amazing days of incredible experience reports and expert talks.

Why are we here? I think it's because we are all trying to answer some questions on the next slide. One second.

I don't see the slides behind me, do you? Actually, could someone come out here and make sure that you guys are seeing what everyone else is seeing?

Just to continue. So why are we here? I think it's because we all share a common belief that agentic coding will reshape technology organizations, just like DevOps did, but 100 times bigger and probably more importantly, 100 times faster. We believe that agentic coding can massively elevate our ability to achieve our organization's goals, to really move the needle on things that matter, and that we believe that there are so many looming questions that every technology leader is facing.

And just to put this into perspective — I was talking to a group CTO of a large private equity firm, and he said something shocking to me. He said, "We are facing the greatest leadership challenge that we've seen in the last 100 years." And that's a big statement. And so I asked him to really justify that claim. And over 15 minutes, I found myself increasingly and very thoroughly convinced. Essentially what he convinced me of is that for at least 60 plus years, leaders could inherit any process in technology, and if it worked well enough, they could take credit for it. And if it didn't work, you could make some minor changes and take credit for that. But now we're entering an era where coding is heading towards instantaneous and free, and so now every process we have is profoundly insufficient — the bottleneck will move into every part of the organization that's not technology. And if the job of the leader is to enable teams, is to eliminate obstacles, leaders are going to have to solve problems that there's no precedent for. And so I found myself absolutely convinced that we are now facing the greatest leadership challenge of perhaps the last 100 years.

And just to cast how people who I so tremendously respect and admire are framing it — this is Adrian Cockcroft. Many of us will recognize his name because he was the cloud architect at Netflix. In 2009, he was brought in to migrate Netflix infrastructure from data centers to running entirely in the Amazon cloud, and that's what set the stage for the Netflix that we know and love today. And he said last summer, "Maybe we should call agentic coding, vibe coding, 'no-dev.'" Now, just to set some context for this — back in 2012, he noted that many people called DevOps "no-ops," meaning that we don't need infrastructure and ops anymore. Of course, that is not true. And so in an age where every developer using AI coding agents is now like a director of development, commanding a fleet of agents, maybe we should call it no-dev — meaning, do we need developers anymore? And so obviously that is not true, just as no-ops was incorrect. I think it shows, one, how mean-spirited we are in the technology community, always throwing functional silos under the bus. But two, it's just showing how we are reshaping technology organizations.

In fact, DevOps really shifted all responsibility left to developers. And I think for a long period of time at that, there was no place left to go. And I think what we're now recognizing is that technology organizations get reshaped, and they're going to be increasingly moved into the business organization that we serve. DevOps reshaped technology organizations in 2009, when we went from deploying once every year to 10 deployments a day, as famously highlighted by John Allspaw and Paul Hammond, the VP of operations and development at Flickr. Suddenly, development had to work differently, QA had to work differently, ops, information security — we all had to shift left. And again, that's going to happen again, but 100 times larger and again, 100 times faster.

So I just want to take a couple of minutes to share with you some key points in my adventure into this strange new age.

This is Steve Yegge, my buddy who I've admired for over 12 years, and I finally got to meet him in June 2024. We had many things in common, but one of the things that we discovered right away is we both have this passion for AI and how it's changing coding from underneath us. One of the things I'll never forget is that we pair programmed together in September of 2024, and we were able to write something that I've wanted to write for over a decade in 47 minutes. And I can promise you, if you've had something that you've wanted to build for 10 years and you can do it in 47 minutes, it will change your life. And it certainly changed mine. And that's many of the things that went into the Vibe Coding book that came out one year ago.

And one of the people that we so much admired was a gentleman named Dr. Eric Meyer. He is considered by many to be the greatest programming language designer of all time. If you've ever used Visual Basic or C# or LINQ or Hack or used Haskell, he has been a fundamental part of those movements. And he said, "We are probably going to be the last generation of developers to write code by hand, so let's have fun doing it." Part of his motivation for that — which showed the inevitability of that — is that time is our most precious resource. Time can neither be stored nor generated. And so if we can find a way to save time, there's almost a moral imperative for us to take it.

And so in 2025, at this conference, one of the most amazing experience reports was from Dr. Topa Bradapall, who is presenting tomorrow. Back then, he was the VP of architecture at Fidelity Investments. He was talking about this: one of the many applications that he had to manage as part of his responsibilities was the software supply chain inventory process, where he could see at any point which of the 20,000 applications there have Log4j. And I think for many years, he's dreamed of what this application should look like. But every time he asked his team to do it, they would say, "We'll need to hire two or six or more developers, and it will take about seven months." And so he was so frustrated — that's my interpretation of it; Topa, you'll have to verify whether this is true — but he was so frustrated that he basically vibe coded for five days straight and built something from scratch that was exactly the application that he wanted. A whole new front-end presentation layer, going directly into the Neo4j database, and that went into production one year ago. And I think it just showed that you can build production applications that are ready for the enterprise — not necessarily customer-facing — but do it in a matter of days, not months.

And so early this year, we decided we can't wait an entire year for another conference. And so, with basically four to six weeks of planning, we announced the Enterprise AI Summit in April in San Jose. And my heartiest thanks to Ben Grinnell, Jeff Galmar, Marguerite Kim, CEO of IQrev, to show that we could pull together a conference and do these conferences more frequently. And that resulted in an amazing conference where we opened up the aperture a little bit. This is a conference that has typically focused on not the tech giants — not the Facebooks, Amazons, Netflixes, Googles, and Microsofts — but large, complex organizations that have been around for decades or centuries. And so we invited organizations like OpenAI, Amazon, Netflix, in addition to a lot of familiar characters.

And it just motivated and substantiated a quote from Amy Willard, one of the technology leaders I admire most, from John Deere. She said, "We all got rebooted at the same time." It doesn't matter whether you're a frontier AI lab, a tech giant, a startup, or a large, complex organization that's been around for 105 years making tractors — we're all figuring it out at the same time.

And just to motivate this a little further: when you talk to people from the frontier AI labs, they'll say, "We've had access to the models for over six months." But that's the technology. If you take a look at the way they're rewiring their organizations, I think at best they have a two-and-a-half months head start, which means that we're all at the starting point at the same time. What a magnificent opportunity to take the lead and close any gaps.

So after Amy Willard spoke, Steve Yegge made comments like, "Wow, John Deere might be ahead of Google" — pointing out the fact that if you have 30,000, 60,000 engineers at Google, the same organizational change adoption curve kicks in. You can't expect them to all change overnight. And there are two ways to interpret his tweet. One is, "Look how great John Deere is." The other way to interpret it is, "Look how behind Google is." This tweet got two and a half million views. One of the replies was from Sir Demis Hassabis, the CEO of Google DeepMind, Nobel laureate, saying, "What an unfair characterization of Google. Tell your pals at Google to go back to doing real work." But I think it just shows how sometimes politically charged this AI-ification of software development is.

So here at this conference, you're going to hear some amazing talks — primarily experience reports. We're going to hear technology leaders talk about their organization: here's our industry, here's how we compete, here's the business problem we set out to solve, here's what we did — including tools and techniques so that you can replicate them — what we learned, what the outcomes were, especially around testimonials, and here are the challenges that still remain. And I love that because this is really a form of the scientific method. You're stating a hypothesis, what you think will happen, and what actually happened. It's a way for all of us to learn as a community.

And so over the next two days, you're going to hear amazing experience reports from almost every industry vertical. I've gone through every one of these presentations. You are in for a real treat, and I suspect that you'll be learning just as much as me.

There's a second category of talks that we'll be showing, which is expert talks — people who are pushing the frontiers of performance. Everything from agentic coding to how to do code reviews, to security risk, and even a peek into the Toyota Production System. I'm so delighted that you're going to hear from some of the experts I admire most, teaching you what they spent a lifetime — or maybe two years — getting really good at, in this world where no one knows anything. As Kent Beck said, "The safest answer to most questions these days is nobody knows."

So to wrap up, I just want to share a couple of things. This is a conference that started in 2014, called the DevOps Enterprise Summit. And I think for those of you who were there, there was a sense of universality — we're all facing the same problems. And I take just so much professional pride that over the years, we've had 1,500 speakers from technology leaders share experience reports across almost every industry vertical.

And one of the ones that I find so delightful is Siemens Healthineers. This is the organization that developed the software for CT scanners — the thousands of CT scanners you find in hospitals. The business problem that they set out to solve was: how do we make it possible for every software build to potentially be certified, to go into the certification process, so that it can be rolled out to the entire fleet of CT scanners? And so through tremendous engineering, they were able to go from six months to certification-ready software to having certification-ready software at the end of each day. Do I think that they're going to be vibe coding software for those mission-critical pieces of equipment today? Probably not. But do I have confidence that sometime in the next couple of years, there will be elements of that software that will be vibe coded using coding agents? Absolutely.

So with that, I just want to share with you our goals. We want to assemble the most switched-on leaders — both team leaders and ICs — who are advancing at the frontier, bringing these amazing technologies into our organizations. We now have a monthly AI forum where, because meeting once a year, meeting even twice a year, is not fast enough.

And I'm going to introduce you to my co-MC, Andy Ramirez, head of growth marketing from GitLab, who challenged me with one of the biggest audacious ideas I've ever heard: how can we enable technology leaders to create communities around them, to enable us to do what DevOps did, but 100 times faster, 100 times bigger? He had this audacious goal to build a community of 1,000 technology leaders within a year. And I can't tell you how shocked I was by that challenge and how bought in I was. And so we're creating this thing called the Software Engineering Leadership Forum, of which you will hear more over the next two days.

So with that, I'd like to introduce my friend and co-MC, Andy Ramirez. Andy.

Andy Ramirez

Thanks. Good morning. Thank you guys all for having us. It's really an honor to be here. It's an honor to share a stage with Gene.

GitLab is extraordinarily proud to also be the premier partner for the first East Coast Enterprise AI Summit.

(Applause)

It is fun to be here, and it is cool to be in a room full of fellow travelers, kindred spirits. There are people here who are working every day to solve for what AI means in enterprise software development. And it's true that you are as smart as the people you hang out with, and GitLab could not be prouder to be part of this room.

I also want to say that it would not be possible to do this without sponsors, and so I want to say thank you on our behalf to Coder, Digital.ai, Equal Experts, Liatrio, Lucid, Delphix by Perforce, PlanDeck, PlanView, Quotient, Veracode, and Exalt. Without you, we couldn't do it, and if you get a moment, please go visit them during breaks or at lunch.

I also want to make sure that you all have access to the full conference agenda. You can scan the QR code, or ask your agent to take a look at it for you and ping you when you need something. And of course, you want to make sure you get on the Wi-Fi. If you'll notice, this is the same Wi-Fi credentials from San Jose, so they should just work. Please also connect to the Slack channel and make sure that you're interfacing throughout the whole experience.

And just one last little bit — thank you, Gene, for everything again, and I'll let you introduce the first speaker. Thank you so much.