In today’s episode of the “I Am A Mainframer” podcast, Steven Dickens sits down with Thomas Amodio
President, Infinity Systems Software Inc., Vicom Infinity, Inc. On this podcast, Tom discusses his journey with the mainframe, advice for those just starting their journey with the Mainframe, and where he sees the Mainframe going in the future.
Steven Dickens: Hello and welcome. My name’s Steven Dickens, and you’re listening to the I’m a Mainframer podcast brought to you by the Open Mainframe Project. The Open Mainframe Project is a Linux Foundation collaborative project focused on the mainframe as a platform, bringing together the open source community around this particular architecture. I’m joined today by really one of my great friends on the platform, Tom Amodio from Vicom. Hey, Tom. Welcome to the show.
Tom Amodio: Thanks, Steve. Great to be here.
Steven Dickens: So Tom, just to get our listeners orientated, could you just really give a brief introduction to your role and what you do at Vicom so it can just get everybody orientated here and just get started?
Tom Amodio: Sure. So I’m one of the owners of Vicom, the president of the company, and I’ve been pretty much a mainframe technology person my entire career. So, I’ll say I lead the company from sales, engineering, and anything to do with the technology itself.
Steven Dickens: And just give us a double-click on Vicom, Tom. I know you guys well and the role you play and some of the leadership roles you play in the industry and the ecosystem, but just maybe expand on Vicom and kind of what you guys do just to help everybody get a picture of the role you play.
Tom Amodio: Sure. Yeah. We’re a little bit of a different reseller. So we’re sort of in that classification of evaluated resellers, but we didn’t start out that way. We started out as a company that was more focused on software sales and systems integration on the mainframe. So if we go back to our origins of Infinity Systems back in the early 90s, our focus was looking at the new stuff coming out, and those things at the time were parallel SYSPlex, was Unix system services, open MVS. TCP/IP was a big movement in the early to mid-90s and into the late 90s. So we spent the first seven years or so as Infinity Systems really enabling these new functions and capabilities on the mainframe. We started to see this movement into value-added reselling, and we lobbied IBM for a number of years.
Tom Amodio: And then by creating this partnership with my friends over at Vicom, we formed Vicom Infinity, which really became a System z reseller as well as focused on peripherals around the z environment, storage, tape, et cetera. But the focus was always on new technology. In our early days, we did some really bleeding edge stuff. In 1998, we created the first-ever Java runtime on OS/390. That was really the; I’ll say, the stimulator for what became known as WebSphere on OS/390 later on. They had WebSphere running on distributed, but after they saw us being able to deliver a Java object on OS/390, it sort of woke them up that the mainframe had a place in this world of Java. And a couple of years later, we actually deployed the first-ever production WebSphere on OS/390 at one of our clients.
Tom Amodio: We did several others after that, some very large-scale ones. I won’t name the customers, but they were very large financial institutions in the New York Metro area. And we really were always experts at everything new. When IBM brought up Linux in Poughkeepsie, two of my engineers were invited up there that day. So it was a couple of engineers from Böblingen and Poughkeepsie, and two of my engineers were in the room. So we could say we were there the first time they did it in Poughkeepsie. And we walked out of there and immediately jumped on that bandwagon. So, once again, our focus has been really on integrating anything new technology that comes out in the marketplace. In the early 2000s, we were doing a lot of CICS web enablement.
Tom Amodio: We did the first-ever CICS web interface, the first-ever Java gateway into CICS. So we’re really on the bleeding edge of technology. We were working closely with Hursley and Santa Teresa, Toronto for UDB, and we would basically look for customers and pilot customers for first-ever implementations on a mainframe for something new coming out of the labs. And that was really our heritage. And that’s what it still is today. I’ll talk later on about some of the new things we’ve been working on in the last few years, but our heritage has always been to break the ice, sort of breakthrough the glass, and find out how we can bring something into the mainframe that up to that point in time was really not really a mainframe technology, but now integrating the mainframe with it and allow it to play in that world.
Steven Dickens: And Tom, that’s certainly how I think IBM looks at you and the rest of the community looks at you. It was kind of taking that leadership role as one of the first adopters of new technology, really diving in at that deep end and saying, “Okay, this new technology is being launched or even before it’s launched, how do we help clients deploy that?” And I know from my engagement with you guys over the last sort of 10 years, or so that’s been a key part of what you’ve been doing in the open-source community, and obviously, a lot of our listeners are focused on that piece. Maybe just, if you could go through some of the examples of what you’ve been doing in the Linux space and some of the open source space with Zowe, that’d maybe bring us a bit more current and talk about some of the newer pieces.
Tom Amodio: Yeah, well, we started our own open source projects back in; I’d say it was around the 2012 timeframe. We just wanted to play around with a few technologies, MongoDB, Nginx. There were a few others that we were playing with, and the bottom line was we wanted to figure out, can we get these open source products to run on z? And we were able to. It actually ended up a conversation I had with some people in Poughkeepsie saying, “We’ve got to do a lot more around this space. If my small little company could get these little products going, what could we do if we have masses working on this?” And we saw a big movement in 2013, ’14, and it only accelerated after that. And we could tell right now the number of open source projects around the mainframe or available on the mainframe; it’s just grown significantly in the last five years. Zowe was a great addition to that in terms of creating a community. It started obviously with Rocket and CA and IBM, but it’s expanding. And the great thing there is, this is, I think, the beginning of what we’ll look back on.
Tom Amodio: It’s still, what I would call, in its infancy, but this is the beginning of really creating an open-source community really to drive new capabilities for the mainframe, solving problems for the mainframe that have existed for years, that we’ve been dependent on vendors to try to solve. But now we’re seeing individuals step forward or even individuals that work in companies that are working in their spare time to contribute. One of the projects we worked on this summer was basically creating a JSON interface for SMF data. SMF data is very valuable. I’ll refer to it as a very valuable resource in terms of mainframe information, what goes on from everything from security through performance through how companies use it for chargeback capacity planning, problem management.
Tom Amodio: There are so many things that go on within SMF data, but being able to unlock it and put it into a format that can easily be ingested into other systems. We have very powerful analytic systems today and systems that can do analysis and graphing, and other things. And the only inhibitor was how to use this 40, 50-year-old format of SMF data and make it ingestible just like any other data into these analysis engines. And that’s was we brought in an intern; we basically created some specifications, mentored him through the project, and on the other end of it, at the completion, we delivered a great component, and the component was, I believe it was just recently adopted by the Open Mainframe Project and the code, the project name was ZEBRA. So you could Google that and look it up. Still, it really is about opening up that very valuable SMF data to systems and platforms and products that can render it and give it a greater, useful life beyond what people use it for on the mainframe today.
Steven Dickens: And Tom, I mean, I’m obviously aware of this, but as we were looking to build the Open Mainframe Project, you were an obvious choice of one of our founding members. I know, and Len has, throughout the years, has been a great servant to the platform, who’s your CTO, as our board chair. Maybe talk us through why you, as the CEO, see it’s important. You’re a relatively small organization. Why do you sort of think it’s so vital for you to take a leadership role in an open source community such as the Open Mainframe Project?
Tom Amodio: Well, if I go back, again, we’ve always been interested in anything new, right? That’s coming out, right? And that could be the technology that started maybe in the distributed world that we think is going to eventually evolve into the enterprise space. And when I say enterprise, I mean including the mainframe, or in some cases, it’s stopped short of the mainframe, and we figure out how to go that extra yard to bring it into the mainframe world. But I think the great thing about the Open Mainframe Project is it builds that community. It gives you access to resources that, before this, the mainframe really was led by a number of large software companies and IBM. And it was every project or anything that was ever done on it, and believe me, I worked for a software company, so I know how this works, and there had to be a dollar value to everything you did, anything you’d developed.
Tom Amodio: There had to be this return on investment. In the early days of Vicom Infinity and Infinity Systems, we didn’t look at; when we said we wanted to figure out how to create a run time for Java on OS/390, it wasn’t because we were thinking, well, we’ve got to make money at this one. We saw that there was going to be a need for it. And we figured if we can solve the problem, the money will come later on. We’ll figure out how to monetize it later. But we led into it just with one thing only, and that was we knew we had this problem. We talked to several customers, and they voiced interest in seeing Java objects run on the mainframe. So that was it. We didn’t have to go out and request a budget and create a business plan and do anything. It was just something I could do myself, and I can fund it, and we can run. And that’s sort of what the Open Mainframe Project allows us to do again, is we can solve problems.
Tom Amodio: We can create opportunities for new things, better integrations. It really advanced the platform without having the drag and the friction of being formal, relying on software companies to solve the problems. This becomes a community working together to solve a problem.
Steven Dickens: And how do you see that? We’re sort of five years into the Open Mainframe Project now. How have you seen that kind of moving forward? What’s been your perspective there, Tom? As you’ve seen from the first day, we launched the project through sort of five years on? How are you feeling that’s going?
Tom Amodio: Well, like anything else, I think it takes time for people to understand what it’s about. Right? So, when you talk about open source, and you talk about the mainframe, there was sort of a gap in terms of understanding. So the people that live in this mainframe world, we were brought up to understand early on, in my days when I worked at Chase Manhattan Bank, we weren’t allowed to just download what I’ll call shareware software, right, the old CBT stuff. They were so afraid of malware and other things entering the environment. So there were people who looked down upon things that we know refer to shareware. Later on, open source took over the distributed world. We saw that happen, and it took some time. But I think to some; the mainframe was still stuck in its old ways of we’re just not going to jump into this open source world right away.
Tom Amodio: So I think it’s been, the beginning days, the early days, the first year or so, might’ve been a little bit slower. We still see some friction in some of the large enterprises in terms of joining the Open Mainframe Project. And I think that’s part of that is they still don’t recognize the value that it can bring in terms of what open source did in the distributed and the cloud world, right, and how it accelerated the cloud. They’re not getting how it can accelerate the mainframe, and it could accelerate solving some of the problems. I mean, customers every day deal with the same problems day in, day out, and they become accustomed to it. It becomes the new norm. And what they don’t realize is that there is a lot easier way to do things. The reason why the cloud is accelerated, and we see distributed systems have grown exponentially is that we’re able to do things with a lot less friction.
Tom Amodio: And I think that’s something that the mainframe sort of still has a problem with, or it’s probably its greatest challenge. So, in terms of the Open Mainframe Project’s growth, I think it could have been better if people understood it and were more supportive of it. I’ll give you one more statement there. And that is, if we look at what’s going on in the mainframe world in the last 25 years, there’s been a real gradual slowing of bringing new skills in, right? It started probably somewhere in the early 90s and continued through the early 2000s and then basically just went off a cliff where you just hadn’t seen in the last 15 to 20 years anybody new really coming in. You see small sort of onesie, twosies. You see what’s going on in the universities where they try to bring new people in, but it’s nowhere near the number for what’s happening on the other side with the attrition, right?
Tom Amodio: So we’ve got this big gap. And I think part of that, when I go back to the problem with the Open Mainframe Project, it is sort of that same issue there where we have just a mentality of leadership within these enterprise spaces that, for some reason, don’t apply open source and open source thinking into the mainframe. They apply it to the cloud. They apply it to their distributed environment, but they sort of stop at the mainframe. And I think that’s something that will shift over time as we bring new people into mainframe management, people that have come from a distributed world and now are learning the value of the mainframe they’ll feel more comfortable. Maybe we’ll see something like the Open Mainframe Project really reach its full potential because I think right now it’s somewhat hampered by large enterprises that are not willing to commit to it completely.
Steven Dickens: It’s always been a big part of that, Tom. I mean, it’s a big part of our project and where we see that collaborative effort within the Open Mainframe Project. You mentioned about skills. What role do you see kind of Zowe playing in that overall space? How do you see it, and kind of how’s it impacting some of those client conversations?
Tom Amodio: Yep. So, Zowe, I think, is still in its infancy, and I’ll say for several reasons, one, in terms of it being adopted broadly, we’re really in the very early nascent stages of Zowe, right? I think people are sort of looking at it. They’re trying to get comfortable with it; understand it. And again, this just goes to the gap there is in mainframe technologists, mainframe systems programmers, what they do on a day-to-day basis and then relating how Zowe can help them get to that next level. So, the greatest thing we could do is have Zowe embedded somewhere in the operating system that, when everything is installed, Zowe is automatically enabled. We remove all the friction from having to go in and enable it as a secondary or an afterthought. Right? So I think so where we’re at with Zowe still is that I want to say somewhat the infancy stage of deployment.
Tom Amodio: The customers, I could tell you just from us playing with it, it’s not simple to enable. And what I mean by that is it should be anybody who has the ability to enter commands and the ability to read a simple installation guide or cheat sheet, get started here type of instruction set, should be able to install a product anywhere. And I think Zowe is not 100% there yet. I think when it does get there, and we’ve figured out how to embed it and make it part of the infrastructure itself, right, where it’s automatically enabled. Then it’ll offer much more promise to the enterprise in terms of then understanding how to plug things into it to enable API and other interfaces. But again, that’s where the community comes in to really build around Zowe, right, and make Zowe interesting. It’s like when the web came out in the very early days of the internet; you just didn’t really know.
Tom Amodio: Okay, this is looking a little weird. I’m going back to the early 90s, but it took three or four years, and then all of a sudden, content and graphics and other things started to really expand, and the internet started to get interesting. So I think that’s sort of where we all with Zowe. We’ve got to get it to the point where it really becomes interesting and has a lot of cool content and a lot of good function and features built into it, or that’s easy to enable with inside it. And it becomes that gateway to everything you need to unlock within the mainframe.
Steven Dickens: Yeah. I mean, I think that’s how I look at it as well, Tom. I think as these projects look to get bootstrapped and get it off the ground, they obviously sort of start to build momentum. And I think if I look at where Zowe’s been over the last couple of years, the first year was just getting its feet underneath itself as a project, getting that first code release. I think that as we start to move through the code releases, we’re starting to see kind of that mature. So it’s interesting that you kind of see it in the same way.
Steven Dickens: Yeah. It’s an evolution. And I mean, that kind of leads me on as we pivot and talk about kind of you and your career; you’ve kind of been with Vicom now awhile? What would be some of that advice you’d be starting to sort of think about giving? We’ve got a lot of younger listeners to the show, early professionals, and I know you do a lot of mentoring with Vicom as part of the Open Mainframe Mentorship Project. What would be the advice you’d maybe be giving to the 22-year-old Tom Amodio as he was embarking out on his career? What would be your kind of advice to your younger self, if you could do that?
Tom Amodio: Advice to my younger self. So I’m going to twist it. If I had to give… So I look at it two ways, right? If I could go back and redo anything, what would I redo or second, knowing what I know now, what would I tell somebody else to do, in a sense? I’ll tell you the one thing that I did in my career that I think is at the core of some of the success I had was I forced myself to learn as much about as many parts of the technology as I could. I’d never felt comfortable just understanding one part of the technology. All right? And if I could tell anybody coming into this platform and then that is you’ve got a career that’s going to span 30 plus years, every year and as often as possible within a year and break it apart as much as you want, but always try to learn and understand some other component. The mainframe has a vast amount of technology built into it.
Tom Amodio: And if you only understand small pieces of it, you’ll never be able to really reach your full potential. The more you understand about how it works and how all the components work within it, the more valuable you are to the enterprise, the greater the value you build for yourself in your career. I never wanted to be somebody that had to be dependent on somebody else telling me how something would work, or I didn’t have to know it to the being the subject matter expert on it. Still, I never felt comfortable not understanding it. Right? Whatever it was, whether it was a database technology, whether it was the networking side, whether it was security, performance, capacity planning, chargeback, I tried to learn as many disciplines as possible around the mainframe. I was fortunate to get to wear a lot of hats.
Tom Amodio: One of the things we always said, even as a company early on, if a customer came to us and said, “Could you do X?”, whatever X was, the answer was never “No.” We’d always figure it out. And that’s what I would tell anybody who’s young starting out this career, and that is continually challenging yourself. And then I’ll give you an analogy. When I was working at CA early on in my career, I didn’t know assembler language. And they hired me on the basis that somebody saw something in the interview, and they said, “This guy can learn anything.” And so I had to learn assembler, and they gave me six weeks to do it. And I said, “I felt like I was at the bottom of Mount Everest looking up, and you can’t even see the summit. I mean, you just could barely see a third of the way up the mountain,” and that’s sort of what it felt like. You couldn’t see where the top was.
Tom Amodio: You didn’t know how long the climb was going to be. And once you scaled it, you realized, “Okay, that was scary, but we did it.” And then, later on in the career, you encounter those over and over again. Right? But what happens is the first time you’re at that base looking up at the summit, it’s scary. After you scale the summit several times, the next challenge you encounter later on in your career, the fear is gone. The trepidation is gone. You just really look at it and say, “Okay, I’ll figure it out. I don’t know what I’m doing today. I don’t know what’s ahead of me. I don’t know how I’m going to get to the top of this, but we’re going to figure it out.” And that’s how we took on every project that we’ve done in the last 20 odd years. I’ll kind of digress on one project we worked on to give you an example.
Tom Amodio: In the early, I’ve instilled this in everybody who works for me, that we just got to challenge ourselves to do things that aren’t easy, right? We’re going to solve problems. When we did Java on OS/390, if you understood the amount of hacking we had to do, it was just bizarre. I mean, I couldn’t even tell you what we did. All I know is we got it to work. And when we delivered the hello world, everybody was hooting and hollering, but we did something else recently, and it started out with a hypothesis. So it was can we talk to the mainframe? Can we talk to the mainframe and actually get a voice response back? And we didn’t know what we didn’t know at the time, right? How to do that, right? How does Alexa for business work, right? How does Alexa work? How does Google Talk work? How does Siri work? And we didn’t understand those technologies, but we wanted to figure it out, and the reason for doing it was we wanted something secure.
Tom Amodio: And we knew that if we had to build something for the mainframe, that would be a natural language processor, it had to be secured, had to be enterprise-ready. And we didn’t know what we didn’t know. But we took it on, and it took us a couple of years, a lot longer than it took us to do Java. It only took us six to eight weeks to do Java. This took us well, the first project only took about three months, but that was just a very crude POC. But we went through multiple MVPs, and we solved different problems along the way, including security and everything else. But the point is, in the end, we challenged ourselves to deliver something. We knew what we wanted it to do and today, what we delivered is the natural language processor that runs on Hyper Protect Services for security, can run on-prem on z, or can run on IBM’s cloud on z. And it’s completely integrated with Watson.
Tom Amodio: And you could basically create a chatbot that talks to an API endpoint, ask a question in a natural language, and get an answer back either in the form of a text, a report, or a voice response. So, again, this started out as a challenge. What don’t we know about doing this? And we didn’t know anything. We didn’t know about Watson. We didn’t know how chatbots worked. We didn’t understand how we could create the API endpoints and have them interact with the chatbots. We had to learn everything. So I go back to if you use that as an example, if you have a young person coming into the mainframe, the thing you got to tell them is there’s a lot that you don’t know, and you can’t be afraid to learn, and you can’t be afraid to challenge yourself. And the mainframe is a little different than any other platform because there are a lot of things that happen within z/OS and z/VM, and the operating systems.
Tom Amodio: This is not like any other platform. You’ve really needed to know about a number of things that go on within this platform to really master it. And everybody should be challenged and say, “I want to master this environment. I want to understand everything about it.” And once you do that, you can do anything, and you’ll have a career set for life because this mainframe, I don’t think in our lifetime or maybe even the next generation’s lifetime, is going away. I just don’t see anything replacing it.
Steven Dickens: And Tom, I mean, that’s a fantastic answer. If I could replay this podcast to anybody who’s embarking out in their career, as we’ve known each other these last few years, I think that intellectual curiosity’s come through, just learning some of the conversations and how you push your organization and sort of embed that in how you guys work and operate. That just comes through and delights your customers. As you’re doing the boring stuff with the customers and you’re deploying these platforms, and you’re doing the mundane, you’re always looking to sort of drive that sort of view to the future. And that kind of leads to my last question. I know you’ve got a fantastic perspective on this platform. As you look ahead, sort of three to five years out, where do you see the mainframe going?
Tom Amodio: Well, I think there are two things that are happening in the enterprise space, right? So depending on the way where you are in the scope and scale of a customer, there are a number of customers that will eventually move off the mainframe, right? Because they just don’t have the skills to manage it. Their business doesn’t demand it anymore—other technology. There was a point in time that there were things the mainframe did that you couldn’t accomplish your business problems or solve your business problems on any other platform, or if you did it, the cost would be a mess. The mainframe was always much more cost-effective, but we’re sort of in that, I’ll say, crossroads right now where there are certain applications that could scale really well on other platforms and are scaling well; we see that every day.
Tom Amodio: But there’s still a number of businesses that have just an immense investment in the mainframe, and to take that investment and try to transform it and put it somewhere else makes no business sense at all. I mean, the economics just fall apart and break down and believe me, we live in the transformation space. We’re masters at it. The economics don’t work for the majority of customers. So that means you’re going to be living with this mainframe for more than three to five years. I see, outside, there’s going to be always the exceptions, onesie, twosies customers that say their business has been able to navigate and move away for whatever. They went to packaged software. They’ve been able to move to cloud-based products, whatever. There’s always going to be those, don’t say, I don’t want to call them one of us because it’s more than one of us, but they’re going to be the minority of customers that can find their way clear and move to another platform or move to the cloud.
Tom Amodio: Majority of mainframe customers, especially the larger ones as we get into thousands of MIPS, once you start crossing that 10,000 MIP barrier, where there are a number of customers, it is extremely difficult to look at getting off the mainframe or doing something and going somewhere else. So I think the move is still going to be to… The problem I think of the future in the next five years; it’s not going to be about what does the mainframe look like? Right? The mainframe is going to look pretty much like what it looks like today. There are going to change. It’s going to evolve. We know that, but in terms of how it processes work, what it does, it’s going to be doing exactly what it does today. And there’s going to be some more function and feature and other things. We’re always going to have those little evolutionary things.
Tom Amodio: But I think the most interesting thing that’s going to happen in the mainframe world, and I think within the enterprise in general, is the introduction of AI. And this is something we actually started working on back in 2000, but maybe we’re ahead of our time. We started thinking about the concept of programming robots to do the work of systems programmers. And we actually developed a whole bunch of code around that just as a proof of concept also. But I think what we’re doing now, we see things out of IBM Research and other areas where they’re focusing on what I’ll say pointed problems within the enterprise and possibly directly within the mainframe environment and how do you use AI to do things, find problems? Again, as the mainframe environment grows, it’s more complex than ever; AI can do things in a fraction of a second compared to how humans could operate.
Tom Amodio: They can make decisions or at least provide a result set to somebody to make a decision much quicker, right, by looking at potentially thousands of data points and analyzing them and then coming to a decision-maker to say, “Okay, here’s your problem.” And I think that’s what we’re going to see as we start moving out, somewhere in the three to five, six, seven-year timeframe, is going to be this expansion of AI into what I’ll call enterprise ops. And specifically, I think it’s how we solve the resource problem for the mainframe, which is everybody’s worried about what do we do? How do we replace the masses that are going to be retiring? Well, we’re not going to be able to replace all of them, but what we’ll do is what we’ve always done. We’ve gotten by with fewer and fewer people. I could remember when I led a systems programming staff of 16 people.
Tom Amodio: In today’s world, you’d only need two or three systems programmers at most to run the type of environment we were running, maybe even less, maybe one and a half, if I really boiled it down, but you need to account for vacation and sick time and other things. But with AI, we really could do a lot more with a lot fewer people at the time. So the question is, how quickly can we speed up AI into the environment and sort of, I’ll say, offset the attrition that’s going to occur in the five to 10-year time frame.
Steven Dickens: Tom, fantastic answer. I was talking to Len Santalucia just before this podcast and said to him, “I love recording this show. It’s a highlight of my week whenever I get to spend time with the guests.” But I was particularly looking forward to this because I think you’re one of the unique people in this community that gives us a view, are not only the business impact but also the technology in your role as the CEO of Vicom. I know you straddle both environments perfectly. So now I just want to say, Tom, thank you very much for being a guest on the show and really enjoyed our conversation here. We’ve run longer than we normally do, but that’s because I think you’ve just got this interesting perspective that we can share with our listeners. So Tom, thank you very much for being on the show.
Tom Amodio: No, my pleasure. Great talking to you, Steve.
Steven Dickens: You’ve been listening to the I’m a Mainframer podcast. My name is Steven Dickens. Thank you for joining us. If you like the show, click and subscribe to the links below. And if you’re on the Apple platform, give us a five-star rating. Thank you very much for being with us on the I’m a Mainframer podcast.