Guilherme Cartier de Palma, Associate Director of Infrastructure/Cloud Architecture at Kyndryl, joined host Steven Dickens to share his insights on mainframe transformation in a recent episode of the “I Am a Mainframer” podcast.
Cartier has over 10 years of experience working with mainframes, beginning his career in SysOps at IBM when he was 18 years old. He believes that a balanced approach is needed for mainframe transformation, with a strong focus on people, processes, and tools. He also emphasizes the importance of understanding the cultural impact of transformation and being prepared to change the way you work when adopting new technologies.
Cartier is optimistic about the future of the mainframe, believing that it will continue to be a core pillar of IT organizations and become even more powerful as it is integrated with the cloud. He highlighted the following key points during his conversation with Steven:
- The importance of understanding the cultural differences between mainframe and non-mainframe teams
- The importance of collecting metrics to track progress in mainframe modernization projects
- The importance of investing in mainframe technology and learning how to manage transformation through daily work
Cartier’s insights provide valuable guidance for organizations that are considering mainframe transformation. His emphasis on the importance of people, processes, and tools, as well as the cultural impact of transformation, is a reminder that mainframe transformation is not just a technical project. It is also a cultural change that requires careful planning and execution.
Watch the episode on YouTube:
Listen or watch the episode on Spotify:
Announcer: This is the I Am a Mainframer podcast, brought to you by The Linux Foundation’s Open Mainframe Project. Episodes explore the careers of mainframe professionals and offer insights into the industry and technology. Now, your host, senior analyst and vice president of sales and business development at Futurum Research, Steven Dickens.
Steven Dickens: Hello and welcome to the I’m a Mainframer podcast. I’m your host, Steven Dickens. I’m looking forward to today’s conversation because I’ve got Cartier de Palma from Kyndryl on the show. Hey Cartier, welcome to the show.
Guilherme Cartier de Palma: Hey, Steven. Hey, thanks for having me.
Steven Dickens: Yeah, thank you for joining us. Really looking forward to this conversation. Big fan of what Kyndryl are doing. Just for maybe the audience’s benefit, I spent about two very productive and happy years in what was Global Technology Services within IBM before it spun out to become Kyndryl, so I’m looking forward to this conversation. So, before we get any further into the discussion here, explain your role, what you do, and we’ll use that as a jumping off point.
Guilherme Cartier de Palma: Excellent. So, I’ve been part of GTS, as you mentioned, from the last 10 years until it spun off as Kyndryl. And in those years, I’ve always been working with the mainframe in different roles, as a SysOps, a prediction support analyst, and most recently I’m working in the global team within Kyndryl for the mainframe division that we call the Core Enterprise and zCloud, and I’m part of the engineering team. So, our team role is to set the strategy and understand the mainframe strategy and the mainframe modernization, and advocate that across our organization and with our customers as well. My role specifically, I am a DevOps architect and a modernization architect as well. So, I talk a lot with customers to understand their struggles and their hurdles that they’re going through, their digital transformation, their modernization journey. And also, help them to achieve a DevSecOps transformation as well.
Steven Dickens: So, my knowledge, and the team at Kyndryl keep me pretty well briefed, my knowledge of where you guys is in the market is probably one of two firms that really get the mainframe. So, there’s a lot of people who outsource mainframes, and I won’t mention their names to maybe protect the somewhat innocent. But those guys, your mess for less. They’ll take onboard your mainframe. They’ll maybe do a decent job of running and operating it, but there’s no innovation there. There’s no driving the mainframe platform forward and really innovating for those clients where they’re looking after the mainframe. As I say, I put you and probably Ensono in a category of two that is really driving the platform forward. You’re not just looking to run and operate it, you’re looking to evolve it. Am I reading the market right? Is that really where you kind of see Kyndryl’s role?
Guilherme Cartier de Palma: Absolutely. Since we spun off from IBM, we’ve been getting a lot of more flexibility in the sense that we can expand more in our line of business, which is actually services for our customers. So, we now are able to partner up with other organizations as well to really create a more cohesive strategy, and really understand the mainframe modernization and where it should be, and how can we really help our customers? So, you are right on the nail with that.
Steven Dickens: So, I mean it’s interesting you mentioned mainframe modernization a couple of times. I think if I asked 10 people what mainframe modernization is, I’d get 11 different answers.
Guilherme Cartier de Palma: Exactly.
Steven Dickens: And I’m probably as guilty of this from my perspective as an analyst commentating on the market. What do you see as mainframe modernization?
Guilherme Cartier de Palma: So, that’s a great question because, as you said, if you ask different people, they will have completely different answers. For us at Kyndryl, mainframe modernization is split into three different categories. Because it’s a very broad subject, so it might mean different things for different people, and that’s natural. So, for us, it can be the modernize on, so taking the mainframe as is and modernizing as best as we can the current platform. The second type of modernization is to integrate with. So, we understand that the mainframe is capable of integration, it’s a highly sophisticated machine and it’s the central of our customers or IT organization, and it’s well-capable of integrating with the remaining of their ecosystem.
So, we can integrate that platform with the cloud platform or any other technology as well. So, that’s the second category. And the third one is the move off or migrate off. Because we understand that not all workloads should run on a mainframe platform. And for some customers, it might make sense for them to migrate some of that workload from outside of the mainframe. So, for us at Kyndryl, we have this holistic point of view on modernization, so we can really help our customers where they want to go, where it makes sense for them.
Steven Dickens: And I think from chatting to Richard and chatting to Petra, that’s a strategic perspective that you’ve got that I think resonates with kind of where the market in reality is. We see some of the hyperscalers maybe banging a drum, which is, “Get off the mainframe. Get to the public cloud.” And whilst that might be the right answer for some small workloads, and I’ve got a pretty pragmatic view, whilst I think some workloads definitely should be on a mainframe, maybe other workloads shouldn’t.
Guilherme Cartier de Palma: Exactly.
Steven Dickens: But I think what I like about the Kyndryl approach is of the three vectors of your strategy, you’re not obsessed with the, “Get off,” also, two thirds of the strategy is innovate and connect, and transform what you’ve got. And I think that’s really positive, and I think really balanced.
Guilherme Cartier de Palma: True. And we have this philosophy of the right workload in the right platform. I mean we know that the mainframe is the most sophisticated platform out there, it’s the most reliable and robust platform and all of those things. So, it’s really fit for mission-critical kind of application. Usually transaction heavy, but not necessarily. We can also run a lot of batch in there, but usually very mission-critical kind of applications. Other kind of applications that are not very critical, it might not make sense for you to run that on a mainframe. So, we can help the customers to move that to a different platform if that fits their scenario.
Steven Dickens: Yeah, I mean my personal perspective is just because the car was invented, doesn’t mean the train, the boat, or a plane is a less valid form of transport. I travel a lot to trade shows and conferences. I drive to the airport and then I jump on a plane, then I might jump in an Uber, which is a sort of consumption-based model, to get to the hotel. And then they might put me on a bus to take me to the conference venue in the morning. I’ve used multiple different types of forms of transport because the journey was different, the consumption model was different, and I think people who are taking that same perspective as they look at workloads, what’s this workload’s characteristics? What should I run it on as a platform? Those are the most evolved companies.
So, this is where my alignment with where Kyndryl kind of makes sense. So, maybe let’s drill down maybe some of the projects you’re working on, some of the things that you’re doing. I’m conscious you probably can’t mention customer names, but just give me some of the examples of some of those modernization journeys that you’re taking clients on.
Guilherme Cartier de Palma: Absolutely. So, it is a very interesting question because the nature of the work that we do in our team, it’s quite variable because we might be meeting with customers to discuss their modernization strategy around DevSecOps. Recently, we did that with a couple of customers, and interesting to notice some of the challenges that we see in their transformation journey. Because when we talk about DevSecOps, traditionally in a technology or organization, the challenge might be the adoption of different processes and different tools. And sometimes, the technology is a challenge. But in the mainframe, not necessarily the technology is the biggest challenge for those organizations because we know that the mainframe can integrate with any other technologies, highly sophisticated as we said, right? So integration wouldn’t be the challenge. It’s definitely part of the workload, but it’s not the biggest one.
The biggest challenge that we see is the cultural aspect, to try to transform the culture. Because although we have some of the brightest minds working in engineering organizations in the mainframe around the globe, in our customers organizations, they are mostly familiar with one very specific way of working, the mainframe way of working. And when we talk about DevSecOps, we talk about all those different tools, all those different processes, and really, a different way of seeing the problems that they solve every day. So, it was very interesting to work with these specific customers to really get into their organization, interview their engineers, and understand the pain points that they’re going through in these transformations, so we can really help them to try to adopt some different practices and collect some metrics that can indicate their progress.
Steven Dickens: I mean it’s interesting. I’ve long said that the mainframe worships different gods and speaks in different tongues. And I think a lot of the challenge that you’re describing is one I’ve seen and observed. It’s not that they’re doing different things, they just describe it in a different way.
Guilherme Cartier de Palma: Exactly.
Steven Dickens: We talk about DASD being connected to the mainframe as just one example. It’s just storage connected to a server, but the mainframers use a different word, which then when you go and talk to the storage team or the SAN team, it’s OSA cards for a network, not a NIC. There’s just subtly different words that mean there’s just a bit of friction in the process. You talk about CICD pipelines in a sort of DevOps world. I don’t know whether the mainframe guys kind of see themselves in a CICD pipeline type of methodology. They are, that’s what they do. Maybe they’re doing it in Devver, maybe they’re doing it in some of the IBM tooling, and maybe the mainframe development team are kind of here. But I start to see some of that changing, and the mainframe vendors are doing a good job of that. You guys I’d throw into that mix. I mean some of those friction points, is that what you see?
Guilherme Cartier de Palma: Exactly. That’s definitely some of the friction points that we see. When you said the mainframe worship different gods, that’s so true because although we do very sophisticated and hardcore complex technical stuff in the mainframe, folks still struggle with some basic concepts for DevSecOps. But at the end, it’s very simple. We just need to help them to see the similarities, the connection points, what they already know. So, one of the examples that we see inside our own organization, so we have those brilliant engineers working with us and they were struggling with basic command line interface commands on Linux. So, those are very basic kind of commands. And I spoke with the person like, “You develop in assembler, you speak a very high-level, absurd technical language and you are capable of doing those amazing things that I could only dream of. You can bend the CPU to your will.” And I mean it’s very simple for you to understand those things as well, but it’s a process, right?
Steven Dickens: Yeah.
Guilherme Cartier de Palma: There’s no rush to the process.
Steven Dickens: Yeah. I’ve got a really good friend of mine who’s jumped from a couple of different organizations, now runs the Linux mainframe team at a big bank. And we’ve had lots of conversations about z/VM. For me, dude, it’s just a hypervisor. It’s just a hypervisor. It’s exactly the same conceptually as what ESX does in the distributed and public cloud space.
Guilherme Cartier de Palma: Exactly.
Steven Dickens: If you know how to bend z/VM to do what you want it to do and be able to configure mini discs in a z/VM environment and do all of that detailed tuning, creating logical partitions in Prism, doing what you need to do in z/VM to sort of carve up a workload, you can get your head around ESX and visa versa. And I think just because the naming conventions and the words and the phrases are different, sometimes we just lose that cross-pollination of the disciplines.
Guilherme Cartier de Palma: Absolutely. Totally.
Steven Dickens: And it makes it harder for vendors like yourselves of, how do you take people on this modernization journey?
Guilherme Cartier de Palma: Yeah. So, again, it is all about the process, right? Really understand where the customer is right now. So, part of our work is to really sit down with the customer, try to map their value stream of their applications, try to understand the point of view of each stakeholder in the process, because the nature of our customers, the mainframe organizations is that they are often very large organizations, and their software ecosystem is too big for a single team to understand what’s going on under the hood. The applications are too complex, they probably touch a lot of things. So, one team might not understand the impact that the changes that they do might impact another team or another part of the technology stack.
So, our work is really to sit down with those guys and map out this value stream. And from that, we can really start this process. So, we talk a lot about the cultural aspect because this is something that we can really not ignore, right? Because it’s going to be one of the biggest challenges that the customer will face. But we also talk a lot about the technology, about the automations that they need to implement in their environment in order to make sure that the transformation move forward.
Steven Dickens: Yeah, it’s interesting listening to this conversation. It seems as much you are on the cultural and the language and the change methodology, you’re less on the technology. Obviously, the technology’s an underlying component of it. Am I reading that right from our conversation?
Guilherme Cartier de Palma: It is because… Actually, I would say it’s very hybrid approach because we cannot do one without the other. So, we really try to make a balanced approach. Because if you only go in the technology side… You probably have interviewed a lot of folks that only talk about the technology side as well, how that’s important. If you go in this approach of the transformation, only talk about technology and automation, “Oh, okay, you need to implement a CICD pipeline.” That’s all true. That’s perfect. That’s all things that we do. But don’t forget about the culture, don’t forget about the transformation that your organization will need to go through. So, we really try to go in a more balanced approach when we talk with customers.
Steven Dickens: I just hold a simple rubric in my mind. People, process, tools. And I think if you can focus on all three equally, you’re probably in a good space.
Guilherme Cartier de Palma: Exactly.
Steven Dickens: So, as we start to think about moving towards the end of the podcast here, we always ask a couple of questions. I think we’ve not spent as much time on your career arc, so maybe I’ll pull that in. Go back to your 22-year-old self. You’re finishing college. Maybe give me a two-part answer, give me a view of your career and then maybe give me a view of what advice you would go back and give yourself if you had chance to go back and speak to your younger self.
Guilherme Cartier de Palma: Right. Around that time, I was already working with mainframes. I was a SysOps, and I was really invested in the technology. So, basically, what I would say to my younger self is, “Continue to invest in this technology because you’re on the right track. The mainframe has a lot of future, so it makes sense to continue learning about the technology, but also don’t forget about how can we transform our organization’s culture through our daily work.”
Steven Dickens: So, maybe sort of fill in some of the gaps. You were starting as sys admin back at 21, 22. Has that time always been at GTS, and now Kyndryl or have you moved moved around?
Guilherme Cartier de Palma: Always GTS. Yeah, I joined IBM GTS when I was 18 years old, and I started working in operations and I’ve been there since then, over the past 10 years. So, I worked as a SysOps, then I worked as a production support analyst. Then, in this global position, I started working as an automation developer, creating global automations that our teams could leverage in our customers’ accounts. And right now, most recently I’ve been working as a modernization architect.
Steven Dickens: Oh, fantastic. So, the other question I ask every guest, and I’d really be interested in your perspective because I think you chat to some fantastic customers in your role. You’ve got a crystal ball, you’ve got the chance to look into that crystal ball. Where do you see the mainframe five years from now? So, we’ve got z16 now, maybe 17 and 18 have come out. Where do you see the mainframe ecosystem in five, seven years out from now?
Guilherme Cartier de Palma: So, from what I see and from what I discussed with customers and from my own opinion, I can really see the mainframe will continue to be the core pillar of our customers organizations, the IT ecosystem. There’s no going back, right? The mainframe is really the central component of our customers’ IT organization, and it will continue to be. But in five years from now, I can see more integration capabilities. So, the mainframe will be much more integrated with the cloud, for example. Customers will be able to leverage much easy their mainframe resources from their cloud side of things. So, the mainframe will no longer be like this very important machine, but it’s sitting in the corner of the data center. It’ll be a much more integral part of their daily activities, and it’ll continue to modernize and it will continue to be very, very important to our customers.
Steven Dickens: I think that’s a great way for us to summarize. Cartier, really enjoyed this conversation. Thank you for joining us on the show.
Guilherme Cartier de Palma: Thank you very much, Steven.
Steven Dickens: You’ve been listening to the I am a Mainframer podcast. If you like what you’ve heard today, please click and subscribe. Click all the buttons, that helps the algorithm and helps us grow the show. We’ll see you next time for another episode of I am a Mainframer, brought to you by The Linux Foundation.
Announcer: Thank you for tuning in to I am a Mainframer. Liked what you heard? Subscribe to get every episode or watch us online at openmainframeproject.org. Until next time, this is the I am a Mainframer podcast, insights for today’s mainframe professional.