Strategic Topics To Think About For 2026, Part 1

February 9, 2026 Philippe Magne

Longevity is an amazing thing. ARCAD Software has now amassed more than 34 years of expertise centered around everything related to DevSecOps transformations. To start off the new year, I would like to share six strategic ideas that will help shape 2026 in the IBM i market, and beyond it.

The big new development for ARCAD is our fourth brand, called DISCOVER, which is focused on application intelligence. It’s the newest addition, and it’s very promising. We’re delighted to have convinced more than 25 customers by year 2025, and I think this will evolve significantly this year.

DISCOVER is all about understanding the existing information system. Every company has an information system, and every company shares one characteristic: the people who revolve around it are not there forever. Anything that centralizes knowledge about the information system and makes it easy and productive to disseminate is necessarily something that resonates with everyone. In that space, AI is obviously the key differentiator, because it is the entry point to achieving far greater efficiency in transferring knowledge. We are combining our proven deterministic technology with AI to provide a highly cost-effective and reliable solution that meets the needs of auditors and IT teams alike. More about our AI strategy later in this article.

Our second brand, DROPS, for multi-platform deployment orchestration, has also evolved a lot and is now positioned in what we call platform engineering. Platform engineering is a deep evolution of DevOps transformation. Specifically, companies have built countless CI chains here and there and realized that their complexity requires a lot of resources. So our focus is on optimizing DevOps resources because, typically, DevOps engineers are not exactly plentiful, and it is better to share them across teams.

Our DOT brand continues its path in data anonymization and subsetting, with a significant contribution from AI to boost our capabilities in detection of personal data while simplifying data anonymization in multiple use cases such as data analytics.

Finally, our historical brand – ARCAD, which shares the company name – has also evolved significantly. We now position ourselves in “Legacy” DevSecOps, not only IBM i DevSecOps. This is simply to illustrate that we are entering an adjacent market to IBM i, namely the mainframe IBM z market. Stay tuned, because a lot is happening on our side right now.

Across these four pillars, we have nearly 500 customers with over 4,500 licenses in 75 countries using ARCAD tools.

This brings me to our true purpose. This is important because “Breaking Barriers,” our slogan, is becoming more and more meaningful and tangible. We are about connecting, typically, Dev and Ops – that is the core principle of the DevOps movement. But also, and above all, I would say, bridging legacy and digital. Our value proposition is that we truly master the legacy world while also having expertise in the digital world. The goal is to bring these two worlds closer – worlds that often tend to stare each other down and engage in trench warfare, which is not beneficial to any company. There is a natural siloing phenomenon that needs to be fought by every possible means, and technology is one of those means – particularly our DROPS tool, which really connects those two worlds.

Another natural gap can exist between applications and infrastructure. That gap is obviously fading over time, especially with the evolution and rise of clouds, and the goal is to achieve far more flexibility across the application – infrastructure boundary.

And let’s not forget the gap between generations. When you come from legacy, you are naturally part of an older generation – like me – and we must also pay attention to Gen Z, which represents our future. The topics on the table are endless, with one central leitmotif – so we will start with that theme.

We are all probably close to overload on this AI topic, but we have to acknowledge there are major movements underway. This is a significant revolution, which is why it gets so much attention, and we will see how 2026 will be structured. Here, I used Gartner’s Hype Cycle. In technology adoption, there is always an initial phase with a peak of expectations. Depending on companies, technologies, and business sectors, you can be on the upward curve or on the downward curve. In general, we are still in the “hot” phase, because we now have some perspective on what AI can or cannot bring. Now we will enter the famous “digestion” phase, to reach a plateau where we have processed it and can truly benefit from these innovations.

I do want to emphasize something that personally bothers me: We tend to forget the downsides. Every revolution also comes with things that are not great. The first is job destruction. You can see it again today, with banks announcing job cuts specifically because of AI integration. We should not kid ourselves: AI is often more job-destructive than job-creating. And there is also the carbon footprint, which is concerning. Climate change will not wait for us, and we need more pragmatism in our approaches. In particular, do not use a sledgehammer to crush a fly.

With some hindsight, as a business leader, I can say that the AI revolution is a blessing in certain ways – among other things, to get teams moving. It triggers self-questioning at all levels, which is very interesting because it can accelerate change. At ARCAD, we are quite used to change.

Some of my colleagues make fun of me because I always say, “This year is a turning point.” For once, it wasn’t me who said it. It was the members of my strategic committee that confirmed that this year, 2026, really is the turning point.

Another interesting thing I have observed is that AI is a unifying topic. Everyone is facing the same revolution, everyone has to position themselves, and everyone is on equal footing in a sense. That’s interesting because it’s not just “experts” on one side and “the people who run the company” on the other. It is deeper: how do we collectively find solutions to digest this revolution as effectively as possible, internally – where we use AI in most of our departments – and externally, where, as a technology company, we must be ahead in how we consume these revolutions.

The first two things you need to ponder for 2026 are interrelated: Pragmatism versus hype with AI, and GenAI in particular, then DevSecOps transformation and AI integration is the second thing. This is as true for ARCAD as it is for any software company, including yours.

The reality with legacy applications is that application knowledge is fragmented. Business rules are embedded in code, not necessarily in documentation. There is a major human risk: Ensuring the transition of skills from experts heading into retirement. This generational transition that we’ve been talking about for the last fifteen years is clearly accelerating. Now we are in the thick of it.

Our vision is based on a shared foundation: A universal baseline for any type of code, including legacy code; a progressive increase in automation. Progressive is the key word, because a big bang does not work, plus a strong dose of security and tests as guardrails.

What the market expects in 2026 is to industrialize your pipelines on critical environments, adopt VS Code as a universal IDE, and increase adoption drastically. That is already a first answer to the generational transition. Young developers do not want to manipulate an old language like RPG or COBOL. What they want is modern tooling.

Now, about generative AI: We need to get out of the fantasy. Honestly, people saying “it’s mind-blowing” is starting to get old. We need a pragmatic approach.

AI changes the game profoundly. It transforms our working methods and opens up clear acceleration capabilities. What is truly interesting is that the development domain is opening up to populations beyond the small group of developers. Now business teams can take back ownership and achieve a form of autonomy in creating new applications.

AI does not replace everything. It is a powerful assistant, but it cannot fully substitute human expertise and understanding.

We are at an interesting phase. With our 34 years of experience, we have historically done deterministic algorithmic work – starting at point A and arriving at point B in a guaranteed way. AI, by contrast, is probabilistic. It explores the space of possibilities and does not always guarantee the same result. The key to success is the combination of the two. These approaches complement each other. Determinism guarantees; AI accelerates and explores.

With that, we need to talk a bit about IBM’s Project Bob. It is a product being launched by IBM and is creating a small revolution within IBM itself. Bob is a development assistant, compatible with IBM i, and AI-assisted.

How do we see Bob? Bob holds much promise, but to ensure the long-term effectiveness of Bob in the IBM i world, we will be combining its AI-first assistance features with ARCAD’s proven, deterministic logic. How? With what we call MCP servers. If you know AI technology, you know this is a standard proposed by Anthropic that generalized very quickly. It allows LLMs to use existing assets rather than starting from scratch every time to search for information. It is a way for LLMs to consume existing knowledge and be more productive.

The clearest illustration is our metadata repository. This is the fruit of our 34 years and our analysis of thousands of applications worldwide, with a depth of information that is, to this day, unmatched. I say that with full awareness. This accumulation of expertise around application analysis is one of the real sources of our technology’s value.

That is what we will provide to a product like Bob – but not only Bob. It can also serve other tools, like Copilot, for example. In any case, it will ensure greater relevance when you use these tools, which have the advantage of being far more user-friendly and more attractive to young developers.

It is always easier to converse with something than to put in place heavy processes. Very soon – by this spring – we will have an MCP server available around the metadata repository, DevOps processes, and also the converter to free-form code. Since 2013, ARCAD’s RPG converter has successfully converted thousands of programs into free-form, is set to strengthen Bob’s features in this area. Another indication that it is the combination of AI and proven deterministic logic that will lead the field. Stay tuned for upcoming ARCAD announcements this quarter. . . .

Philippe Magne is chairman and chief executive officer of ARCAD Software.

This content is sponsored by ARCAD Software.

This Issue Sponsored By

Table of Contents

Content archive

Recent Posts

Subscribe

Pages

Search