Accelerating Mainframe Modernization with IBM’s Aparna Sharma

This is Textron tv. Hey guys, thanks for the throw. We're here with Aparna Sharma and we're talking about IBM's acquisition of an outfit called Advance because they help modernize applications.

And Par NNA is the, see if I get this right, managing partner for hybrid cloud services for IBM Consulting. It's a mouthful right there. Partner and welcome to show.

Hello. So good to be here and good to be talking to you. People talk about modernization of applications all the time, but it really means a lot of different things for a lot of different folks.

So maybe you could walk us through everything from one of the options that people are doing. 'cause it seems like sometimes people will just refresh a GUI and call it modernization, and other times they'll refactor the entire application. It's a spectrum of things.

How do we know which one to do when? Alright, that's a very good question. And, um, um, I think, yes, of course there are many options, many definitions, uh, for different enterprises.

But, uh, from my point of view, I think, um, it's not just modifying the gui, right? Uh, depending upon what the, uh, application and if it's a core application, uh, unless you modernizing the core of the application, you're probably not gonna get the benefits. Right.

Um, now in terms of what does modernization mean and what are the various options, you know, it could be, it could be refactoring as you said, it could be retiring in application, if you would. A lot of these applications have grown. There are, you know, there's duplication and functionality and such.

Um, it could also be, um, you know, moving, uh, to public cloud. It could be moving to hybrid cloud. Um, and, and the core part of this is knowing what to do when, uh, and we believe in hybrid.

We believe in fit for purpose, and there's a lot of resonance right now. Hybrid cloud is here. So figuring out, you know, if no two applications are the same, every application, every workload is different, and, uh, fit for purpose provides the best value.

Um, then, uh, how do you analyze what goes where, what goes in public cloud? What remains in on-prem? What remains in hybrid?

And what's the methodology that you're going on? And that's what is really modernizing of applications. Of course, IBM has been at modernization of applications for a long time, especially in the mainframe era.

How does this acquisition move that ball forward? Yeah. You know, this continues to be a very big challenge for our customers, Mike.

Yeah. Um, every day when I'm talking to customers, there are multiple challenges that are coming up, right? Uh, first one that I hear about all the time is, uh, is business agility, right?

So if you think about, you know, everybody is driving business transformation and, uh, mainframe applications and data are a very core part of it. So how do I bring in the same development velocity, same development quality, right? With my mainframe applications as well.

Second thing that I hear about, um, is risk mitigation. So by definition, a lot of these applications are core mission critical applications. So by virtue of that, the risk is high.

But then, um, there is also knowledge continuity risk, right? So a lot of people who used to possess this knowledge, you know, either are retiring or, you know, uh, going away from the talent pool, right? So enterprises are figuring out how am I gonna support these applications if I don't have this, uh, this knowledge?

And that's why further need for modernization, if you would. Um, and the third one, of course is always cost, right? So, um, over a period of time, my workloads have grown, my applications have grown, my, you know, entire landscape has proliferated.

So there is a need to rationalize and there's a need to, um, you know, optimize. So I hear that from customers, you know, all the time. And, um, um, and it's a very important problem because, um, you know, mainframe continues to be a very integral part of core applications for large, uh, enterprises, right?

Um, and that's why, um, we are strengthening and it's very exciting. We are strengthening our capability, uh, through this acquisition. Uh, this is a carve out.

This focuses purely on application modernization, especially on, um, mainframe. And, um, they do all kinds of things. They've, they've been doing this for over 30 years.

They do a, a lot of, um, a lot of things around assessment, automated factoring, database migration optimization, archiving, automated testing. And there are several combinations of this as well, right? So, you know, are you going from COBOL to Java?

Are you going from COBOL to modern cobol? Are you, you know, you have some probably exotic technologies in your back, natural or aaba or ca n or, and so on. So there is a lot of capability and a lot of scenarios, right?

And they carry a lot of expertise, uh, in all of these areas. And that's why we are so excited. So essentially we're strengthening and widening our capability.

There's a lot of excitement these days about all things generative ai, especially when it comes to creating code and maybe even converting code. Is this whole effort going to get simpler? 'cause I feel like a lot of folks feel, well, if I engage in this, it's gonna take two or three years for us to accomplish this task.

And by the time we get done, we're not sure what the outcome is. So can we shrink the amount of time required? Oh, absolutely.

Um, so I think if you think about like, if you take one use case, right? Um, there are about, you know, based on a survey that we did, there are about 230 plus billion lines of cobol, right? That are estimated to be actively running by enterprises.

So you are absolutely right. Um, you know, this problem can't be solved by hand. This problem can only be solved by, um, automation.

And, um, um, I don't know if you had a chance to see it. Um, last year in Q4, uh, we released, uh, a solution called Watson Code Assistance for Z. And this is basically a solution that helps modernize mainframe co applications with capability that include application discovery, auto refactoring to business services, transformation of COBOL code to Java, but really using generative AI and autogenerated tests as well to validate the quality of the new Java code, right?

So you're not getting jal, the quality of the Java code is, um, is good, and we very excited about that. Uh, we are doing several pilots as we speak and proving it in the client and one solution to really be able to do this modernization otherwise by hand, think about millions or billions of lines of code. It's just not possible.

So we are really leveraging that gen AI to bring in that automation, to bring in that acceleration and more importantly, reducing talent risk as well. So you've been at this a little while. What do you see organizations doing when they start these projects that just makes you shake your head and go, folks, we need to be smarter than this Need.

Yeah, no, absolutely. So I think, you know, different for different customers, the, the journeys are different, and I can give you a lot of examples, but there are several strategies that our clients are leveraging to modernize, right? So, so first we talked a little bit about leveraging the clout, uh, the innovation on the cloud, um, as well as all the investment that they made on the mainframe, uh, over the years, right?

And leveraging, leveraging that, um, investment, if you would. And they're leveraging this hybrid model across their estate, and the ones who are doing that are able to drive, you know, much, much higher business acceleration as well as driving better. ROI.

Now we see common transformation patterns across DevOps, microservices, cloud native, hybrid cloud integration, API ation, as well as large scale transformation, right? Um, and so, you know, a Japanese financial services company, right? We modernize their applications to microservices architecture.

We, um, uh, we help them modernize on a cloud-based architecture, hybrid cloud architecture using cloud native application development. They integrated the cloud native applications with their ZOS mainstream services. Um, and this results, this approach really resulted in 70% faster response time, right?

Um, and uh, another example that comes to mind is a global retailer. So they establish high availability, low latency coexistence between the current current systems and applications. And, uh, real time capture of data change in the new system was really essential for them to, um, uh, to minimize, uh, any impact to the current online and batch processes.

Or there's an insurance company, very large insurance company here in the United States, and they, they have the vision to be customers first and best choice in products and services that it provides. And to achieve this vision, this company is innovat on all platforms including Z. And to be able to enable, you know, fast and productive DevOps across the enterprise, not just across mainframe, not just across, um, distributed, but have a single DevOps, a single CICD, uh, if you would.

And, um, they're targeting efficiency. They're expanding their skill resources across all of these platforms. And this is really accelerating, you know, release of, you know, the, just their development velocity release of new code in, um, in production and also simplifying access for their business critical applications to core systems and and services, right?

So the many, many patterns we go through, uh, a very deliberate process with our customers. And, um, uh, we ask them, uh, jointly, we figure out, okay, so, um, what are the challenges? Uh, what are the modernization objectives?

What's the functional mapping of applications and data? What are the technical skill gaps? What are the current or future non-functional requirements?

What are some of the risk and compliance, uh, challenges? And then we go through a discovery process, right? And this is an automated process, uh, where we do a discovery feasibility analysis, both from, you know, a technical functional standpoint, but also from a financial standpoint.

And then we figured out, okay, so what are the various paths for them? Uh, how do they do in place modernization or how do they do cloud native? Or how do they do hybrid cloud modernization, uh, based on what's their, um, you know, cloud, um, cloud provider of choice.

So we, we do all of that, and then we come up with the strategy, uh, cemented by a business case, if you would, and then, uh, we help them, uh, implement that. And we do that in a, um, in incremental fashion, you know, going on the days, Mike, when it would be multi-year projects like big bang modernization. So why your strategy could be multi-year, right?

You do that in shorter sprints in increments. And, um, uh, and we have hundreds of those projects going on, uh, and uh, we are so excited about this new acquisition because, you know, apart from the assets we have, uh, we get access to their assets, their expertise, and we widen that base. Hmm.

So lemme ask you something. I think historically people thought about app modernization as an event, and, and normally you would probably go hire somebody to take on that task because it wasn't something the IT department needed to regularly run. Has that changed though?

Are we moving to an era where at modernization is essentially a continuous process and it's just something I have to bake into my workflows? It is a continuous process, absolutely. And I, and I think that has changed.

Now, sometimes, of course, they, um, they would bring in experts like the, like us, right? To be able to do that strategy, to be able to do that business case, to be able to, um, you know, help them implement, uh, that entire strategy. But at the same time, it's an, it's an ongoing process.

And, um, and, uh, this has visibility, um, you know, across the chain in our client's organization, right? Uh, because, uh, while executives are prioritizing strategic customer centric outcomes, right? If you think about improving customer engagement or accelerating time to market, or reducing their capital expense or increasing revenue or whatever have you, but if their systems are not up to it, right?

They're not able to achieve, you know, any of that, right? So this is, um, a C level, uh, mandate or a c-level challenge. Um, now there is work to be done.

So we did a survey and we can share, you know, some of that offline with you, but, um, while a lot of clients are focused, right? Four and five executives said that they need to accelerate digital transformation. Uh, and they talked about an almost 80% of these of them said that they are modernizing their processes, their, uh, applications and their data.

But within that, you know, there is only 27% of the executives who said that they've been able to complete their modernization, right? Um, now that there are several challenges that they face their financial challenges, technical challenges, expertise, challenges, you know, sometimes even cultural challenges. And that's why they're partnering with partners like us to be able to address that.

So what's your best advice to folks? I mean, as they kinda look at this whole thing, the number of applications I could probably modernize, or the list is large. So how do I prioritize, how do I prioritize that in a way that is, you know, makes the whole thing feasible?

Yeah, no, I think that's the, that's the whole value, right? Uh, by, um, by taking an iterative approach to modernization, right? Which is of course, while we have assets and expertise and so on, it's custom to every client environment.

Um, and the process that I talked about, right? In terms of client challenges, objectives, uh, the discovery process, the feasibility analysis, and then don't make an emotional decision. Make a business decision, right?

Wherever the ROI is the highest, right? And, um, and then of course bake in automation as part of it. Um, you know, whether it's tools for source code discovery or tools for discovery and visualization of data or, um, um, you know, do that, uh, with, um, uh, with the business case.

So we create that roadmap and the business case and we co-create this journey roadmap, right? Um, so clients are able to make, um, uh, make a decision now, um, our advices get started, right? There are multiple ways to get started, and as I said, you know, there are, um, it's, uh, while the strategy could be multi-year, right?

But there are, um, you know, several things that you can do, you know, within a few days. Like for example, we run a mainframe application modernization hypothesis workshop, right? Which takes like two days to one week to kind of establish what's the mainframe context, what the pain points are, what are the potential approaches, and then we have a lot of exemplars, right?

Um, of modernization initiatives. Um, then second, like, okay, you know, we have, uh, a garage based methodology, um, and we say, okay, now if you're gonna rewrite, um, a cloud native application and you're gonna, um, or you're gonna write a new cloud native application and you wanna integrate that with your mainframe, you know, that's a proof of value, proof of concept that we can do and we can hash it out in two, three weeks, right? Um, kind of using our, um, you know, cloud native, uh, development capability, our API expertise, design thinking, agile methods and so on.

So you can get started with DevSecOps, proof of value. We can do that in two, three weeks again, right? Um, we can do this whole entire assessment and blueprinting, usually it takes about six to 12 weeks, and now we are accelerating that with all of these, um, all of these tools and assets and, uh, Watson code assist and so on.

So, so, um, a lot of customers start with strategy, but if they have a strategy already, then it's a question of how do you execute it? And do you wanna do API application first? Do you wanna do DevSecOps first?

Do you wanna develop, uh, an application and then integrate it? So there's several patterns and based on those patterns, right? Um, and the ROI, we help them prioritize their entire journey.

All right, folks, you heard it here from the very minute an application is deployed, it starts to age, and of course, technology's advanced. So in that regard, application modernization, it's not just a job, it's an adventure 'cause it's something we always need to do. Aparna, thanks for being on the show.

Thank you so much. Was so much fun talking to you. All right, and back to you guys in the studio.