Banking on Information with Jasper Geurts, Managing Director Americas at SIG

In this episode of Banking on Information, Rutger van Faassen interviews Jasper Geurts, Managing Director Americas at SIG. They discuss the complexities of software development, the importance of quality assurance, and how organizations can prepare for the future of technology. Jasper emphasizes the need for standardization in software development and the role of GenAI in shaping the future of coding. The conversation highlights the challenges and opportunities in modern software systems, particularly in the banking sector.

Benchmark report
https://www.softwareimprovementgroup.com/finance-signals-2025-report?utm_source=Infobanker&utm_medium=Rutger&utm_campaign=us

Takeaways
- Jasper Geurts is passionate about simplifying complex software systems.
- Software development has become increasingly complex due to legacy technologies.
- Quality in software is defined by international standards that are hard to operationalize.
- Organizations need to benchmark themselves against industry standards to improve.
- GenAI is changing the landscape of software development and coding.
- It's essential to maintain human oversight in AI-driven coding processes.
- Improving software quality involves understanding and remediating technical debt.
- The future of software may involve more automated coding with AI assistance.
- Organizations should focus on cultural and procedural changes to adapt to new technologies.
- SIG provides tools and methodologies to help organizations improve their software quality.

Keywords
software development, technology, SIG, complexity, future-proof, quality assurance, GenAI, banking, standards, innovation

Chapters
00:00 Introduction to SIG and Jasper's Passion
03:01 Understanding Software Complexity and Legacy Systems
06:01 The Importance of Software Quality Standards
08:59 Future of Software Development and AI
11:52 Preparing for the Future: Benchmarking and Standards

Transcript

Rutger van Faassen (00:01.138)
Hello and welcome to another episode of Banking on Information. Today my guest is Jasper Geurts, who is the Managing Director Americas at SIG. Welcome to the podcast, Jasper.

Jasper Geurts (00:13.058)
Thank you so much. Thanks for having me. Excited to have this conversation.

Rutger van Faassen (00:17.871)
Yeah, well, let's dive right in and start with that very important first question. Why, Jasper, do you do what you do?

Jasper Geurts (00:25.87)
So I work for SIG, the company that you just mentioned, and it stands for Software Improvement Group. So I love improving things. I love making complex things simple. And I think the world of technology that I operate and that I work in has become incredibly complex. So Marc Andreessen said it in, I think, a post in the Wall Street Journal or an essay or whatever it was in 2011.

Rutger van Faassen (00:38.055)
Mm-hmm.

Jasper Geurts (00:55.276)
Software is eating the world. And I think software ate the world. Every company right now is in some way, or form a software company. That means that the traditional banks that used to have tellers and a lot of people, now have a lot of software developers and people that build software systems. And these software systems are actually complex by nature. And within that complexity, you have different

shapes, different forms. So a lot of the banks here in the United States, I think globally, are still running dated technology, like legacy technology, such as COBOL. COBOL is a wonderful language, but it's incredibly verbose. It's very old. Like I've looked at certain software systems within my career in this company where some of the code predates me, right? It was built in the seventies and I was born in the eighties.

Rutger van Faassen (01:48.613)
right yep

Jasper Geurts (01:52.748)
So this is a lot of complexity and a lot of old stuff that we need to modernize at a certain point. And that is something that is inherently complex. And at the end of the spectrum, you got modern software like GenAI is taking up the world as well. It's taking the world by storm and everybody is talking about it. I think everybody is using it and a lot of organizations are investing in it and to build systems and GenAI systems to make things more efficient.

Rutger van Faassen (02:14.299)
Mm-hmm.

Jasper Geurts (02:22.414)
When we look at some of the characteristics of these systems, whether it's old legacy stuff or the most modern AI systems, we see common patterns, patterns of complexity and patterns of systems that aren't necessarily performing as well as they could. This is why we are here, why my organization is here and why I do what I do, is I like to help people and organizations.

to improve that, to make software better, to remediate technical debt and those complexities.

Rutger van Faassen (02:58.139)
Yeah. So you like to make things better. You like to make the complex simple, which is actually very hard, right? Making something simple is actually very hard, but it seems like you're very passionate about it and say, let's take all that complexity and let's, let's drill it down and let's make it more easy to understand and then make it better.

Jasper Geurts (03:17.172)
It's the hardest thing. the hard, It always is the hardest thing. Like summarizing, making things simple. Sometimes we write reports, advisory reports for customers and it's easy to write a hundred pager. It's very hard to write a one pager summary of that. Right. So yeah, making complex things simple is very hard.

Rutger van Faassen (03:31.93)
Yep. Yep.

Rutger van Faassen (03:36.899)
Yeah, But it's clear that you're passionate about that. so what does that Explain to me what SIG does? What do you typically do with your customers?

Jasper Geurts (03:46.926)
So what we've done is we created measurement models to make these complexities that we just talked about visible. And this is about source code analysis. Basically, the programming and the instructions that are given by humans and now co-created by Gen.ai, of course, the instructions to computers to execute whatever they need to do.

There are many different ways on how to build software systems. And yeah, let me ask you a question. When was the last time that you were in an elevator?

Rutger van Faassen (04:17.029)
Mm-hmm.

Rutger van Faassen (04:24.709)
Probably yesterday, right? Being in New York. You're always in an elevator, yeah.

Jasper Geurts (04:26.722)
Yes, in New York, so you go to, So what goes through your mind when you step into an elevator? Are you at all afraid that it will come crashing down?

Rutger van Faassen (04:39.797)
No, I don't think about that usually. Otherwise, I might not get in. But yes.

Jasper Geurts (04:43.438)
Not a lot of people do, because most buildings in New York, have elevators, right? And you don't even think about it anymore. But of course, an elevator, it's a dangerous thing. And this is the reason why there's a lot of standardization. You can't just build an elevator company and just start making elevators and installing them in these buildings. And so these are norms. And these are norms that need to be checked. My elevator in my building, it's being checked every quarter. And then you see

Rutger van Faassen (05:04.559)
Yep. Yep. Yep.

Rutger van Faassen (05:12.207)
Yep. The little paper. Yep.

Jasper Geurts (05:12.94)
little signature of the person doing it. They're like, all right, this has been checked. And this is just a highly standardized and governed sector. And software isn't in a very weird way. It isn't. These standards exist though. So there are international standards that prescribe what good quality software should look like. But it's pretty hard to operationalize this. And this is what my company does.

Rutger van Faassen (05:27.686)
Mm-hmm. Right. Mm-hmm.

Jasper Geurts (05:42.318)
We help define what quality means by working with international standard organizations. And we provide technology and methodologies for organizations to work with these standards to ensure that they're building software that is future proof instead of software that just works today, but is harder to change in five or 10 years.

Rutger van Faassen (06:05.199)
That's very good. Now, when you work with your customers, how do they usually say they get value from working with you? What feedback did I give you?

Jasper Geurts (06:16.206)
So we have, with this methodology, a couple of specific products. We have a software product that just runs on all the source code of our customers, and it provides them telemetry and metrics, things that the development teams and the leaders within the organization can use in order to define these guardrails and make sure that everything falls between

what you set, the agreed upon quality metrics. And that's what they really appreciate about working with us because we apply this in a non-toxic way. So you could say every developer needs to write it exactly in this way, and then you take away this creativity from a person. So what we instead do is we say, you're working in the banking sector. That means you're probably working with a technology that can be COBOL but it could also be Java, just a second different thing. And we say, well,

Rutger van Faassen (06:43.182)
Yep. Yep.

Rutger van Faassen (06:57.337)
Bye.

Rutger van Faassen (07:03.492)
Mm-hmm.

Rutger van Faassen (07:07.704)
Yeah.

Jasper Geurts (07:10.7)
Typically people write it in this way and you're doing it differently in that they provide a recommendation on how to improve a development organization. So that's a feedback that we get is it's, easier to work with than, setting very strict standards. And it provides this clarity and a recommendation on how to improve. That's what, that's what my customers provide us as feedback. of this.

Rutger van Faassen (07:13.549)
Yeah. Yep.

Rutger van Faassen (07:33.368)
Yeah, I can see that that's very valuable, sort of knowing where you stand and kind of like relative also to an industry and relative to a standard, because then you can also determine kind of like, is this good or is this bad? And can it get better? Right? Or what do I need to do to make it better? Now you mentioned something interesting, you sort of said being future-proof. So I love to do this thing called Futures Thinking. We don't know what the future is going to be,

but we can think about a possible future. So I'd love to get your thoughts. If you think 10 years out, so 2035, what would be a possible future that you can envision and what does code look like? What does software look like? Give me your best shot at what a possible future could look like.

Jasper Geurts (08:16.78)
Yeah, there's so many possibilities there to answer this. And I really love this question. So what we already see is in this world where people are building software systems and writing code, you see that that is already changing. So you've got co-creation with Copilot and those type of tools that help you create code. And this will most certainly

Rutger van Faassen (08:19.99)
Mmm, yep.

Jasper Geurts (08:44.864)
evolve very soon. But 10 years out, that's a long time. And so will these software systems evolve even further? Yes, they will. Will they start riding themselves? Probably. Hopefully, we're still in control as human beings. I really do believe that. But I think it's going to be incredibly interesting if you see that there are inefficiencies in the way that we

program and we can improve that. that is, yeah, That's definitely going to change when Gen.AI gets more prominent in these processes. yeah, what I also think is that it will create much more code than what we have currently today. And I like to think that that provides an opportunity for my company to contribute to this because

Rutger van Faassen (09:23.692)
Yeah.

Jasper Geurts (09:40.396)
Yeah, if we need less people to do a similar amount of work and the amount of code and the amount of the stuff that you actually need to manage increases, or that you need this type of control and governance around it.

Rutger van Faassen (09:56.535)
Yeah. So maybe less hands on keyboards, more co-pilots doing most of the coding, maybe the humans in the loop still thinking about it, but probably also doing a lot of checking, right? Because that is something, especially now, while Gen.ai is by no means perfect, you do need to check what comes out of it. So in that sense, think what you're doing at SIG is very important and probably becomes even more important as things get written by Gen.ai.

Jasper Geurts (10:23.126)
Yeah, absolutely. I think you hit it on the head there. it's just a different abstraction of complexity, right? So a programming language is a way for us in human language to give instructions to a computer. And Gen.AI is just a different abstraction of that. You can explain it in more natural language and tell it what to do. But you still need to check whether it did what you expected it to do. And again, like this whole thing about I need to be able to change it.

Rutger van Faassen (10:41.226)
Yeah.

Jasper Geurts (10:53.72)
Whether a GenAI or other type of AI helps you with that, that's fine. But I believe that we want to keep, you know, stay in control as humans. So, exactly.

Rutger van Faassen (11:02.016)
Yeah, stay in the driver's seat. Yeah, so what do people need to do to get ready? So that possible future that you painted, what can people do today to get ready to be future proof?

Jasper Geurts (11:14.22)
Yeah, that's a great question. so it's never too late. A lot of organizations, people are thinking about this stuff, certainly in cybersecurity. We want to protect the data of our customers. So there's a lot of governance around that. But yeah, so we released a benchmark report specifically for financial services. And what we believe is that it makes sense to benchmark yourself with your competition. So what are banks currently doing?

Rutger van Faassen (11:22.966)
Mm-hmm.

Jasper Geurts (11:44.104)
And how are systems being built? And if you're lacking behind the industry, what should you do now in order to improve that? And don't get me wrong, it's about billions of lines of code that a bank can have. And that's a lot. You can't change this overnight. And this is why you want to change procedures, culture, and all of these things. So yeah, it's not something simple. This is why we are also here. We are here to help.

Rutger van Faassen (12:11.423)
Yep. Yep.

Jasper Geurts (12:12.814)
make that complexity and complexity more simple than it currently is.

Rutger van Faassen (12:16.194)
Yeah. So benchmark yourself, make sure that you understand where you are relative to others. right? That's probably a good thing because if you only look internally at your own stuff, you never know how you compare against the industry. right? So that's probably a very good first step. And then obviously I think SIG can help once you've made that sort of inventory to identify, where are you furthest away from the benchmark and what are the things that you should prioritize?

Jasper Geurts (12:43.776)
Absolutely. Yeah. Yeah. And it's, it's, it's not just benchmarking is using these standards. think that analogy that I just mentioned about the, elevator, something like this, I really think that people need to think about it like that. It's there. are a lot of these standards and sometimes it feels a little bit dated and like restrictive to talk about governance and all of these things, but they're there to make sure that your future business can, can progress. Right. It's a big problem.

Rutger van Faassen (13:09.908)
Yeah, absolutely. Yeah, well, maybe that's a that's a good point to wrap this up. Thank you so much, Jasper for thinking through the Futures Thinking exercise and for being on the podcast.

Jasper Geurts (13:23.682)
Yeah, thanks for having me.

Rutger van Faassen (13:25.78)
Perfect. Until next time, choose to be curious.