Infrastructure Matters Podcast: Strategies for Successful Mainframe Modernization

Steven Dickens: Hello, welcome to another episode of Infrastructure Matters. I’m your host, Steven Dickens, and today we’ve got Scott Silk from Astadia. Scott, we’ve been trying to get this video on the books for a while. It’s fantastic to finally have you on the show. Let’s introduce you to the listeners and viewers here. Tell people a little little bit about yourself and what you do for Astadia.

Scott Silk: Great. Steven, it’s been a while. I’ve been working on this for six months, so it’s a pleasure to be with you today. I’m Scott Silk.

Steven Dickens: Good things come to those who wait, right?

Scott Silk: That’s what they say. That’s what they say. So I’m chairman and CEO of Astadia, and we were acquired by Amdocs actually in November of 2023. So I’m now running that business as part of Amdocs and on a bigger and larger scale. But my background, I started out as a mainframe guy, worked for Unisys for many, many years, and then crossed over to the dark side and ran startups after that. So my background’s a combination of small company and large company.

Steven Dickens: So you mentioned Astadia. Obviously we’re here to talk a little bit about that. Can you just give me the 30 second overview? Obviously some great news with Amdocs of late. Can you just give us the sort of 30 second overview and then we’ll dive into what we’re seeing in the mainframe modernization market in general?

Scott Silk: Sure. Thanks, Steven. So I joined Astadia nine years ago, and it was based on a hunch a little bit actually. I sold mainframes out of college about a hundred years ago. And so I knew of mainframes. And when I interviewed with Astadia, I had a hunch that, you know what? The death of the mainframe wasn’t mini computers or client server, maybe it’s baby boomers retiring. People my age retiring that’s going to drive people off the mainframe. So nine years ago we set out a strategy to build the ultimate mainframe modernization and migration company, in anticipation for the years 2022, ’23, ’24, when baby boomers started retiring in earnest, and we wanted to intercept that demand. So proud to say we were the largest but also the fastest growing player in the space before we were acquired by Amdocs at the end of 2023.

Steven Dickens: So you talked a little bit about it there. This is obviously a dynamic market. The platform’s just celebrated its 60th anniversary. We’re seeing there’s a demographics component to this market as well, so you’ve obviously got the technology trends and the underlying infrastructure and the code, but then you’ve got that demographics piece also playing in. How have you seen the market maybe over the last five years? You mentioned nine years ago. If you could characterize that five year period, what’s your perspective?

Scott Silk: Yeah, great question. It’s been slowly increasing over the last five years, up to the point where literally 90 days ago we saw a huge spike in the market where our pipeline almost quadrupled when it comes to large deals like Fortune 100 deals. And when we talked to CIOs, we said, “Why is this happening?” And they basically said that you could kick the can down the road migrating your mainframe only for so long, and now you’re at the point where staffs are retiring, our number one partner who I’ve never met, Broadcom, is doing a lot of things to make our ROI more compelling. And now they’re at a point where the risk of staying on the mainframe exceeds the risk of moving. And that’s why we’re seeing such acceleration in the market. It’s really, the risk is shifting.

Steven Dickens: And do you see that profile changing? Do you see the dynamic changing? You mentioned 90 days. Is that part of a broader trend, or is it just literally the accelerator pedal has been pushed?

Scott Silk: If I had to be a prognostic, I would say that that acceleration is going to continue for the next five years. It just seems like we’re really at the front end of a sea change. It’s going to continue to accelerate for the next five years. So that’s an exciting place to be.

Steven Dickens: For sure. And I mean, obviously with the Amdocs piece, that gives you the ability to capitalize on it. One of the big questions that I seem to wrestle with, there’s modernize on the platform, there’s replatform, or there’s refactor. Those are the three big broad choices. You’ve obviously got the BMCs and the Broadcoms, and the IBMs, and others in the ecosystem as well, pushing that sort of re-modernize on the platform. There’s what used to be the assets from OpenText and Micro Focus, that are now at Rocket with that re-platform. And then there’s the refactoring, modernize, move it to a more microservices containerized sort of architecture. That’s how I look at the three buckets in the market. Where do you see it, and do you see demand more in one versus the other?

Scott Silk: Yeah, it’s a great question. So our technology supports what customers want to do. If they want to modernize our mainframe, check, if they want to move to distributed x86, check, if they want to go to the cloud, that’s great. So we can support a hybrid cloud model. But the thing that we learned was quite interesting. We had been a Micro Focus partner for over 35 years before we got into the refactoring business. So we were the leaders in re-platforming services based on Micro Focus and our own technology if it was Unisys. When we got into the refactoring business, our tooling is so automated that we actually… Let me back up a second. When we launched our refactoring tooling, we thought the demand would be split 50/50 between re-platform and refactor.

But the tooling is so automated and we do automated testing, that we can move people to Java or C# to the cloud, faster, cheaper, and at less risk than a re-platform. So we anticipated a 50/50 split between our refactoring and re-platforming solutions. It’s really like 98% to 2%. Because when we present a refactoring proposal and a platforming proposal, by side, the customers asked, “What am I missing here? If I can get to Java or C#, better, faster, cheaper, I’m going to refactor.” And that’s why our business is really growing.

Steven Dickens: Well, probably the toughest question I’ll ask you, so get ready. One of the refactoring challenges I hear consistently is it’s not real Java once you’ve done the COBOL conversion. This JOBOL phrase that’s out there in the industry, it’s a janky migration to something that’s not really Java on the back end. What’s your response to that and Astadia’s response more widely?

Scott Silk: Yeah, that question comes up quite a bit. And when we look at the Java and the C# that we generate, I probably use… Is it going to look the same as if the world’s best Java developer developed it? No. No. Is it more than good enough? Yes. So what we tell our clients, is let’s move to Java, let’s get to the cloud as soon as possible and start recognizing that 60, 70, 80% cost savings. And then let’s go back and say of the estate that we modernized… we’ve migrated rather, what percentage of those applications need to be containerized? Microservices and so forth. And what we’re finding, it may only be 10 or 15%, but the good news is they’re funding that 10 or 15% with house money because they’re already saving 60 to 70 to 80%, because now they’re running on the cloud. So it becomes a self-funding type scenario. A lot of these applications have been chugging along in the back office for 20, 30, 40 years. There’s no need to modernize them. But for the 10 or 15% that’s customer facing, we would help with the modernization of those.

Steven Dickens: You mentioned 60, 70, 80% there. Is that what you are seeing as the ROI in the case? I mean, if you can share customer examples, great, if there’s… maybe they need to be abstracted to protect the names of the innocent. That would also work. But is there any sort of examples you can share of where you’ve seen those ROI type numbers?

Scott Silk: Yeah, on average, it’s 60 to 80%. And unfortunately a lot of the clients we win are Fortune 500, if not Fortune 100. If I mention their name, I get a call from their lawyer five minutes after this interview. But 60 to 80 is what we feel very comfortable sharing with our clients. The highest we’ve done are 92, 93%. We have two clients that have broken 90%. One is a very large federal client and one is a large financial services client. But we don’t want to overcommit, we feel very comfortable with 60 to 80% cost savings. And if we can get you up into the low 90s, all the better.

Steven Dickens: So Scott, we’re 10 minutes in, it’s 2024, and it would be rude of me not to ask you a generative AI question in this year. Every podcast I do has got a generative AI question in there somewhere. Where’s Astadia leveraging it? I’m assuming for code discovery, but can you maybe just drill down and talk about really where it’s embedded in the solution, and then how that’s transformed the discussion?

Scott Silk: Yeah, we don’t talk to any customer within the first five minutes, it’s the GenAI story. And our GenAI strategy is 180 degrees opposite of most of our competitors. And if you think about it, a lot of our competitors are 60 to 70% automated, when migrating COBOL to Java, and they use GenAI to close that gap and get it to as close as 100% as possible. As you know, when we move COBOL to Java, we’re 99.9962% automated, we measured that accurately, and we do automated testing. So-

Steven Dickens: I want to know what that tiny percentage is, at some point, that isn’t automated, but maybe not on this podcast.

Scott Silk: 99.9962, that sticks in my head. It wouldn’t be worth it to use GenAI to get that number up to 100%. So our strategy is 180 degrees opposite of our competitors, because we’re looking at use cases to complement our platform that’s already highly automated. So we’re looking at co-discovery, one we’re building out right now. Another one is modernization. A third one might be automated test case development, things of that nature. So we feel that we’ve got a pretty dramatic lead over our competitors at this point, and we’re going to be using GenAI to maintain, if not enhance that market lead.

Steven Dickens: And has that changed the conversation? We’re obviously, what are we, 18 months, two years into the GenAI story. Has that changed the discussion that you are having with the clients?

Scott Silk: Yeah, it’s interesting. When GenAI came out, everybody said, “Oh my God, we’re going to be obsolete.” And then I started looking at some of our friendly competitors and they said, “We’re going to have what Astadia has.” And I go, “Well, when?” “Well, four to five years.” And I’m like, “Well, four to five years is an eternity.” right?

Steven Dickens: Mm-hmm.

Scott Silk: So what we see is there’s a huge GenAI buzz, and it was a big hype cycle. Now it’s dovetailing back a little bit when reality sets in. But we do believe that GenAI is for real and it’s going to have a big impact on our space down the road, but it’s probably going to be three, four years down the road, rather than six months down the road.

Steven Dickens: So I’ve been tracking you guys for a while, we’ve been chatting on and off getting briefings. Huge moment in the company’s history recently with the acquisition of Amdocs. Tell me what that means. We’ve talked about it off camera, but tell me and the listeners what that means from an investment point of view, a reach point of view, a scale point of view. Because from my point of view, the way I look at it is putting that wood behind the arrow of your solution, is going to be impactful. But am I thinking about it in the right way?

Scott Silk: Yeah, I think so. We’ve gotten into a situation where we were the largest and the fastest growing player in the space, and we started winning deals, 30 million, 40 million, 50 million. Those were really good-sized deals for a company our size. But when you’re dealing with the Fortune 500, if not Fortune 100, they really wanted the backing of a $5 billion company. And we have that now with Amdocs. The second thing that Amdocs brings to the market is dominance in telco and financial services. Two verticals where there’s just a lot of mainframes. And we are a horizontal company. We sell in all verticals, but we see a lot of opportunity in financial services and telco, where Amdocs is dominant. And then finally, you get a kick out of this, when we move somebody’s workload to the cloud, we earn trusted advisor status over time. So once they’re in the cloud, the next question, “Do you do managed services? Do you do DevSecOps? What are you going to do for modernization, predictive analytics?”.

And we say, “Well, we have a partner for that, we have a partner for that, we have a partner for that.” Now we’re vertically integrated, so we can go soup to nuts, all with Amdocs. That doesn’t say we still don’t partner with all our close partners like the Cognizants and the Capgeminis and the Accenture’s of the world. But for clients that want vertical integration, one throat to choke for the whole project, that’s where Amdocs brings capability as well. And we joked that for every dollar we’d earn moving people to the cloud, we probably would leave three to $5 on the table for their digital transformation strategy. With Amdocs, we can now participate in that.

Steven Dickens: So it gives you both the reach into those two key verticals, financial services and telco, but just as maybe the sort of beachhead for a wider Amdocs conversation as well.

Scott Silk: Correct. Correct. And they’ve also done wonders for our R&D budget and our go-to-market budget. So that’s pretty cool as well.

Steven Dickens: So Scott, CIOs and CTOs, they’re wrestling with this game of three-dimensional chess, it’s, “Do I move off the platform, do I modernize on the platform, do I refactor, do I replatform?” This is typically the heart and lungs of an operation, they’re thinking, “I’ve got to get that data under control. It underpins a whole piece of my system of engagement as well with clients.” If you were to boil this down, what would those three key takeaways be, and the three big tenants, if you will, of your conversations with those senior leaders?

Scott Silk: Yeah, so it’s a big decision when you’re unhooking the heart and lung machine and you really have to get it right. So our strategy is all based on risk mitigation, and that’s what CIOs are looking for. So first and foremost, we’re 100% focused on mainframe migrations, that’s all we do. The mainframe is a complex beast, so we really need to live at seven by 24. And unfortunately in some cases we do live at seven by 24. Secondly, it’s all about experience. 28-year-old MBAs are great for some things, but not necessarily the mainframe. So we are looking for people with 20, 30 years mainframe experience. And you always recognize an Astadia consultant when they show up, because they can be seen as the no hair group, the gray hair group, or may use hair products group. So we have an experienced bunch.

And then finally, our tech, both for refactoring and replatforming, is the most scalable in the industry. So we like to say we scale higher, we perform better, we do it at less risk, and we do it more cost effectively than anybody else. So when you add up those three things, the technology, the experienced people, and the focus, we boast a 97% project success rate based on reference ability, dating back to 1994. And then we always say tongue in cheek that it’s a big decision for CIOs, and when they’re successful in migrating their mainframe, they’re going to get called into the CEO’s office for accolades. Unfortunately, the opposite is also true, and when that happens, the conversation’s a bit different.

Steven Dickens: Well, Scott, it’s fantastic to finally get you on camera and talk about this. We’ve been talking for a while. It’s a fascinating story for me. Every time we talk, I learn something, and I hope the listeners and viewers really enjoyed this conversation. You’ve been watching Infrastructure Matters, as I’m your host, Steven Dickens. Please click and subscribe and do all those things that help the algorithm, and we’ll see you next time for another episode. Thanks so much for watching.