Mendix is dead in the age of AI, or isn’t it?
That’s it. The dream is over. Or isn’t it? In this article I share the strenghts and weaknesses of Mendix. I will not be deleting Mendix Studio Pro yet, but over the next couple of years a lot will most likely change, as Mendix is currently doing.
Estimated reading time: 3 minutes
Why would anyone use Mendix anymore when AI can build applications for me?
We’re entering a world where business applications are created entirely from prompts. Describe what you want, press Enter, and watch software magically appear. Which is great news. We can finally stop worrying about all those things that used to slow us down and focus on what’s important: generating more applications.
Need something that connects to six different systems, supports thousands of users, and becomes business-critical within three weeks? Just use AI.
Integration? Just put it in the prompt.
Security review? Prompt.
Multiple environments? Prompt.
Support model? Prompt.
Ownership and governance? Prompt.
Regular platform updates, security patches, evolving best practices, compliance requirements, lifecycle management, scalability, and long-term maintainability? I’m sure there’s a prompt for that too.
That doesn’t sound right, does it? The funny thing is that this isn’t completely hypothetical.
Is it a prototype or enterprise ready solution?
A few weeks ago, I was having this exact conversation with my husband. He’s usually the first person I know to dive head-first into new technology. If there’s a new AI tool, framework, or trend, there’s a good chance he’s already experimented with it before the rest of us have even heard about it.
So naturally, he was explaining how AI would soon be building more and better software.
I nodded along and then started asking some annoying questions.
For the first time in the conversation, he paused. Not because AI couldn’t help with those things, but because he realised those questions are fundamentally different from building an application. They’re the questions that come after the application has been built. That’s because building an application has never really been the hard part. Building something that survives contact with a real enterprise is.
In my experience, most business stakeholders (like my husband) care far more about solving business problems than choosing the right deployment strategy or planning a framework upgrade.
That’s not a criticism. It’s exactly why organisations have architects, security teams, platform engineers, and operations specialists. These people spend years developing expertise in the areas that determine whether an application remains secure, reliable, and supportable once it reaches production.
It’s relatively easy for someone to prompt an impressive application into existence. What’s much harder is understanding everything that sits beneath the surface and makes that application enterprise-ready.
The right context is crucial
When AI becomes capable of implementing all of these concepts automatically, someone still needs to understand that they matter and provide the right context. After all, AI can only optimise for the requirements it’s given. The challenge isn’t generating an application. The challenge is knowing what an enterprise-grade application should look like in the first place.
Mendix was never just about making development faster. Its real value has always been providing a foundation that is maintained, secured, governed, and continuously updated. Security best practices evolve. Infrastructure evolves. Compliance requirements evolve. The platform evolves with them.
What makes this powerful is that organisations don’t need teams of specialists focused on every individual aspect of software delivery. Instead of assembling architects, security experts, platform engineers, infrastructure specialists, and compliance teams for every application, a Mendix team can focus primarily on delivering business value while building on a platform that already incorporates many of these principles.
That doesn’t eliminate the need for expertise, but it significantly reduces the complexity of delivering and maintaining applications at scale.
Mendix also isn’t sitting on the sidelines watching AI happen. It’s investing heavily in it. AI-assisted development is rapidly becoming part of the platform experience, helping developers create applications faster through natural language interactions The goal is to combine AI-powered productivity with the governance, security, and reliability that enterprises still need.
In other words, yes, you can vibe-code. But you can vibe-code inside an environment that’s designed for applications that need to survive beyond the first demo, the first release, and the first excited LinkedIn post.
AI Agents
The same story applies to AI agents.
Right now, every vendor seems to be announcing agent platforms, agent frameworks, agent builders, agent orchestration layers, and probably agent agents. Everyone wants AI integrated into their business processes but turning that ambition into something useful often requires highly specialised skills.
Mendix is taking a different approach by making agent development a low-code capability. Instead of building everything from scratch, developers can integrate agents directly into applications and workflows using tools they’re already familiar with.
Mendix has always excelled at connecting and orchestrating multiple systems. Most enterprises don’t operate from a single source of truth. Their data is spread across ERP systems, CRMs, document repositories, custom applications, databases, and countless other platforms.
An AI agent is only as valuable as the context it has access to. By sitting at the centre of these business processes and integrations, Mendix allows agents to interact with data and functionality across multiple systems, rather than being limited to a single application or data source. That’s where the real value starts to emerge. Not just AI that can answer questions, but AI that can actually participate in and automate enterprise workflows.
The result is that organisations can adopt AI faster, without needing an army of AI engineers or rebuilding their existing technology landscape.
So, is Mendix dying because of AI? If your definition of software development is generating a prototype over coffee, then maybe. If your definition is delivering, operating, securing, governing, maintaining, and evolving enterprise applications for years, then the answer looks very different.
So no, I’m not deleting Studio Pro just yet.
We’re keeping a close eye on all the developments with Mendix and AI and we’ll keep you updated about it.
If in the mean time you have any questions or comments, feel free to reach out to us!
