Join host Steven Dickens in this episode of I am a Mainframer featuring Jakub, a Broadcom professional and chair of the Zowe Technical Steering Committee. Jakub discusses his work advancing Zowe, a major open source project helping developers, system operators, and enterprises integrate z/OS with modern tools, identity systems, and workflows.
He explains how Zowe’s components—including Zowe Explorer for VS Code and IntelliJ, the Zowe CLI, the API Mediation Layer, virtual desktop capabilities, and the Zowe MCP Server—are making mainframe development more accessible and enabling AI tools and agents to interact with the mainframe more securely.
The conversation explores how Zowe grew from vendor-supported beginnings into a broader community project, with customers and service providers increasingly contributing code. Jakub also discusses the project’s future, including AI, Zowe Version 4, improved setup, and the importance of giving developers hands-on access to z/OS.
Jakub shares his unexpected journey from software engineering and the space industry into the mainframe ecosystem, along with advice for early-career professionals: stay curious and invest in lasting relationships. Looking ahead, he envisions the mainframe as a highly optimized part of enterprise infrastructure, continuing to support critical transaction processing while its data helps power AI-driven innovation.
Watch full episode here:
OR
Transcript:
This is the Mainframe Connect Podcast, brought to you by the Linux Foundation’s Open Mainframe Project.
Sponsored by Phoenix Software International.
Mainframe Connect includes the I am a Mainframer series, the Mainframe Voices series, and other content exploring relevant topics with mainframe professionals and offering insights into the industry and technology.
Today’s episode is another in the I am a Mainframer series, exploring the career journeys of mainframe professionals.
Steven:
“Hello and welcome. You’re joining us here on another episode of the I am a Mainframer Podcast. I’m your host, Steven Dickens, and I’m joined by Jakub. Hey, Jakub, welcome to the show.”
Jakub:
“Hey, Steven, great to be here.”
Steven:
“So let’s get the listeners and viewers oriented with a little bit about you and what you do, and then we’ll use that as a jumping-off point.”
Jakub:
“So, hey everyone. I’m now working primarily at Broadcom. It’s the mainframe vendor that all of you have heard about. But apart from that, I’m heavily active in the open source community. That mainly means that I’m running the Zowe TSC committee, which effectively means that I’m kind of responsible for Zowe. That usually manifests in me standing in their presentations at shows and conferences and saying, ‘Hey, if you hate anything about Zowe, I’m the right one to shout at.’ Sometimes people shout at me.”
Steven:
“One person said, ‘Not one throat to choke.’ That’s the way I like to describe it.”
Jakub:
“So that’s basically where I spend most of my time. Apart from that, I’m involved with some of the AI initiatives around Zowe, like the Zowe MCP Server, and the use of Zowe and agents for development within Zowe itself.”
Steven:
“I remember being in the room when we launched Zowe as a project, and the community was a little bit spooked by the fact that the three vendors didn’t own the roadmap and that the roadmap was now owned by the community. I think that was probably the first time—or not probably, it was the first time—that had ever happened for us in the Zowe solution space. That was over five years ago now, which is kind of crazy.”
Jakub:
“How have you done for seven years at this stage?”
Steven:
“Yes, it’s ridiculous how fast that has gone.”
Steven:
“So I imagine most of our listeners know what Zowe is, but could you give us a 30-second or two-minute overview—the value proposition and the pitch, if you will, of what Zowe is and why people should be interested?”
Jakub:
“You should be interested in Zowe if you want to make sure that it’s easy to integrate the z/OS platform with whatever else you have in your enterprise.”
Steven:
“That’s one sentence—you just dropped the mic there, Jakub. But no, expand on what that means.”
Jakub:
“Zowe itself is a relatively sizable open source project with various components. There are components that help developers be more productive within their IDE when accessing and working with the mainframe. This includes Zowe Explorer for VS Code or IntelliJ.
We have a CLI, which is useful for developers, system operators, or basically anyone who automates anything on the mainframe.
And then there is the server-side part, where we primarily have the API Mediation Layer. That manages access to z/OS APIs. It’s relevant, for example, if you want to integrate z/OS with your OIDC provider, or with an external identity provider or enterprise identity system.
And then there is a virtual desktop where you have applications that you can access within your browser in a safe manner.
So this is what it consists of. The new addition—and the one you are working on—is the MCP Server, which is built on top of Zowe, and allows agents and AI tools to interact with the mainframe in a secure and safe way.”
Steven:
“So you’ve described it as a relatively large project. I’m going to try to remember some numbers, so keep me honest here. There are about 1,500 projects within the Linux Foundation, including CNCF, OpenTelemetry, Hyperledger, and all the other major projects. The last time I saw the statistics, Zowe was in the top 30. Is that a bad estimate?”
Jakub:
“I think it’s somewhere around there. We’re routinely among approximately the top 1,000 open source projects in the world. Being in the top 30 in the Linux Foundation is roughly where we fit. I don’t know the exact numbers right now, but it’s definitely not in the 100s.”
Steven:
“So what’s driving that? I think that segment may surprise people. There are still people surprised that there’s open source on the mainframe, which is crazy that we’re still having that conversation seven years on. They may also be surprised that the project is so active, with so many code commits and so much velocity. Why do you think this project has gained that much traction?”
Jakub:
“The original driver was definitely the vendors that were really willing to invest in Zowe. That’s the reality. But as of last year, that has gradually started to change, and we’re seeing more and more users and members of the user community joining us—not just from the vendor side, but also from users.
There is real value in the key message we are sending: z/OS is not an island. It’s not going to be an island anymore. We need to integrate it with the rest of the continent. How do we do that? Zowe is one of the options and answers. There is a real problem, and Zowe is solving at least some parts of it effectively.”
Steven:
“So can you talk about that community engagement? I remember the three vendors that got Zowe started—Rocket Software, Broadcom, and IBM. Can you speak to that community engagement and what you’re seeing as you lead the TSC?”
Jakub:
“First of all, it really depends on whether we’re talking about the part that is more client-side oriented, such as VS Code or IntelliJ. We even have some contributors from customers who use the tools—not just contributing issues, but contributing directly to the code, fixing things, and improving them. And it’s growing. This space is really growing. I think we’re nearing the point where there will be more contributions from other people than from the original vendors. The other part is the server side, which is a bit more complex. The whole z/OS world is far more secretive and less willing to share anything. Nevertheless, even there we’re seeing engagement. I would argue that it’s more from service providers who are using it and helping end users. We also have some direct users involved, but not as much—maybe around 10% of what’s happening within the VS Code or CLI projects.”
Steven:
“I think that speaks to the diversity. If you had said 10 years ago that there was an open source project for z/OS where end users would be contributing code, it would have blown the minds of a lot of people in the community. The fact that we’re now talking about it on a podcast more than five years after the project started is kind of crazy.”
Jakub:
“It’s still not the same as with non-z/OS products. I’ve been involved in other open source projects before, and the variety of communities there is larger. But let’s face it: z/OS is part of critical infrastructure, and when you’re working within critical infrastructure, there are many limitations on what you can and cannot do.”
Steven:
“As we start to look ahead, what should we expect from the directional roadmap for Zowe? Not what’s coming in the next release or the next three months, but what does the vision look like over the next 12, 18, or 24 months?”
Jakub:
“The biggest thing that’s definitely coming and will play a massive role is AI. The Zowe MCP Server is a clear winner in this space. We will also be releasing Version 4 during this timeframe. The plan is to benefit from all the simplifications we’ve made through Version 3, so hopefully it will be much easier to set up on z/OS. That applies not just to the z/OS side, but also to the SSH-based replacement for the z/OSMF APIs that we’re building. This will allow it to run more easily, potentially alongside z/OSMF. It’s not necessarily intended to replace z/OSMF; it’s about being able to run effectively with something we can change as needed.”
Steven:
“One of the things I’d love to get your perspective on is that you have an open source lens into the project and the mainframe community as a whole. What are you seeing more broadly? We’ve talked about AI and talent trends. I see open source as a way to bring new people into the platform and perhaps change the demographics. What signals and trends are you seeing?”
Jakub:
“There are a few things. I won’t start with AI—that’s the obvious one—but I’ll get back to it. One thing that still makes it difficult to join the z/OS world is how hard it is to get hands-on experience. I hope this changes soon. There has been a lot of effort from the Open Mainframe Project to provide access to a platform that everyone can use to validate things on z/OS. I think this will be a huge game changer. Communities that need to test on z/OS would finally be able to do so. We could get more open source onto z/OS, involve more people in the platform, and give them basic experience with it.
When engineers encounter z/OS in an enterprise, they’ll already understand how the platform works, how to access it, and how to integrate with it. Most large enterprises have z/OS somewhere, even if most of their employees don’t know about it. Making z/OS easier to integrate with enterprise identity systems—so people don’t need separate usernames and passwords just for z/OS—would also make things much easier.
You mentioned open source, and the obvious answer is AI. AI is making it far cheaper to develop things, so research and development costs for new functionality are going down significantly. That’s affecting open source communities more and more—and not necessarily in good ways. The group of people who can review contributions and decide what is worth merging into a codebase is relatively stable and isn’t growing quickly. But the number of people using multiple agents to create changes for open source repositories is increasing.Quite often, the results are good. Nevertheless, the capacity to review them is limited.”
Steven:
“That is going to be a challenge.”
Steven:
“So, Jakub, we haven’t got your story arc into the show yet. The show’s called I am a Mainframer. How did you get into the mainframe? How did you join Broadcom? What was your journey to get where you are?”
Jakub:
“Totally random. My career arc would look like some kind of strange zigzag to most people. I always wanted to be a software engineer. Since I was five, I knew I was interested in it—I wrote my first program when I was six. I returned to programming in high school, and after university I studied engineering and information technology.I started with a standard job in a very distributed environment. I spent a few years doing standard Java development for various midsized companies in the Czech Republic. There was nothing particularly interesting about it.
Then I moved into Earth observation and started working in the space industry. Within a year or so, I became head of software engineering at a small Czech company. My role shifted from writing code to building a team that could deliver the code. At first, I hated it. I wasn’t a manager—or at least I wasn’t back then.
After about five years, I became exhausted and decided it was time to move on. I looked around to see who had something interesting to offer in Prague. That coincided with the time when CA started expanding its mainframe teams there.When I heard the word ‘mainframe,’ I only remembered it from a university lecture. But I thought, ‘Okay, let’s see. It may be fun.’ They said it was a project that had received awards for DevOps, so it seemed realistic, modern, and related to what I had been doing.
I joined as a software engineer about six and a half years ago, when Zowe was still a relatively young project. Then, about four years ago, I moved into product work, and around the same time I started running the Zowe TSC. Since then, I’ve been responsible for helping ensure the success of the TSC and Zowe at the governance and policy levels, as well as helping make sure everything is secure and working properly.”
Steven:
“I always love hearing people’s stories—how they randomly get into the mainframe because they’re looking for a change. Fast-forward six and a half years, and you’re running the technical steering committee for a top open source project, probably the flagship open source project on the mainframe. It’s a great story.
One of the questions I always ask on the show is this: if you had the opportunity to go back and speak to your younger self, what would you say? You’re kind of mid-career now, so what would you say to your 21- or 22-year-old self just leaving college? We’ve got a lot of younger listeners to the show.”
Jakub:
“One thing I always say is: stay curious. It’s going to be useful to you regardless of what you’re doing.The second thing is to take better care of your relationships. You’re probably overemphasizing work and work-related matters, and missing the importance of being part of the communities around you. Make sure you build and maintain lasting relationships for the rest of your life. When things go really badly, those relationships are what will help you get through them.”
Steven:
“Great advice. I couldn’t agree more. Your network is a source of strength. Is there an example—people you’ve stayed in touch with for 10 years who have come back into your life?”
Jakub:
“I still keep in touch with my friends from high school. It’s been about 20 years, and we meet every December 20th somewhere in Prague or the Czech Republic and spend time together.
“They’ve been there for me through some very difficult times, including health challenges. You may remove this part, but I went through cancer about three years ago. My friends were there for me and helped me get through it and recover.
Running the Zowe TSC during that time was one of the more unusual experiences. I had been doing it for a year and a half while going through chemotherapy, trying to figure out how to make everything work together. It did, and the community was really supportive.”
Steven:
“If you put the effort into a community, the community will support you when you need it.”
Steven:
“So Jakub, one final question before we wrap. You’ve got a really interesting perspective on the mainframe. Where do you see the platform five years from now? You get to look into the crystal ball, the mist clears, and you get a vision. Where do you see it 5 yrs from now?”
Jakub:
“I see it as an optimized block within the overall enterprise workload that runs a very specific subset of workloads—nothing else. It will primarily handle transaction processing for core processes and the data associated with them. I don’t think it will run much beyond that, but it will remain a critical part of ensuring stability for processes that require extremely high availability. That will be its primary role. Obviously, the data from the mainframe will be consumed by AI, and the information will be used by AI. I would also argue that much of this will be driven by the fact that AI will be doing most of the work around engineering and operations five years from now.”
Steven:
“Yeah, I would agree with that. Well, Jakub, it’s been fantastic to have you on the show. I really enjoyed the conversation. It’s fascinating to have an open source conversation on the mainframe and hear about the latest developments in the project. You’ve been watching us here on the I am a Mainframer Podcast. Please click like, subscribe, share with a friend, and do all those things. We’ll see you next time. Thank you very much for watching.”
Narration:
“Thank you for tuning in to the Mainframe Connect Podcast and this episode in the I am a Mainframer series, sponsored by Phoenix Software International. Like what you heard, subscribe to get every episode or watch us online at openmainframeproject.org. Until next time, this is the Mainframe Connect Podcast.”
#Mainframe #OpenMainframeProject #LinuxFoundation #MainframeConnect #IAmAMainframer #OpenSource #MainframeInnovation #ZOs #Zowe
