Zijn we niet afhankelijk van jullie als dit eenmaal gebouwd is?

Nee. De software is volledig van jou — standaard code die elk gekwalificeerd development team kan overnemen. Je bent niet afhankelijk van Ciphix. Stop je met het onderhoudspakket, dan blijft de applicatie gewoon draaien.

Hoe voorkomen we dat het project uitloopt?

Elk traject begint met Sprint 0 — scope en prioriteiten worden vastgesteld voordat de bouw begint. De iteratieve aanpak zorgt ervoor dat je continu werkende software ziet en op elk moment kunt bijsturen.

Is AI-gegenereerde code betrouwbaar genoeg voor bedrijfskritische software?

Kwaliteit is geen kwestie van vertrouwen — het is een kwestie van controle. Alle output wordt automatisch gecontroleerd op kwaliteit en veiligheid voordat het productie bereikt. De snelheid zit in het elimineren van repetitief werk, niet in het overslaan van kwaliteitsstappen.

We werken al met platformen. Moeten we die vervangen?

Nee. AAD vervangt je bestaande IT-landschap niet — het vult het aan. Bestaande systemen en platformen blijven intact. Welke applicaties baat hebben bij AAD, bepaal je zelf — wij helpen die vraag scherp te stellen.

Hoe weten we dat we het juiste bouwen?

Door te beginnen met Sprint 0. Niet met technologie, maar met het probleem. Wat wil je bereiken? Is maatwerk de juiste keuze of is er een betere weg? Pas als die vragen beantwoord zijn, beginnen we met bouwen.

Whitepaper: from AI experiment to production

How to make the move that most miss

You're not the only one who gets stuck

Somewhere in your organization lies an AI pilot that never made it to production. Or maybe several. They worked fine in the demo. The enthusiasm was there. But as soon as the question came up about when this will actually go live, things fell silent.

This is not an exception. It is the rule.

The technology works. That’s not the problem. The problem is in the step after that: how do you make sure what you build actually becomes part of your organization? Secure. Scalable. Integrated with the systems that have kept your operation running for years.

For mid-sized organizations of 100 to 2,000 employees, this is especially challenging. You have the complexity of a large enterprise – legacy systems, multiple departments, governance requirements – but not the budget and team of a multinational corporation.

The organizations that do move from experimentation to production do one thing differently: They don’t start with building. They start with the right questions.

Build faster without concessions
Develop production-ready applications in weeks instead of months.
Avoid loose AI experiments
Work from a single foundation for integrations, governance and architecture.
Keep control of your software
Build without platform lock-in and maintain full flexibility.

AI works fine. But that's just where the problem begins.

Vibe coding – building applications based on simple AI instructions – has all but eliminated the barrier to software creation. In days you build what used to take weeks.

But what is built quickly is rarely built to last.

Teams build their own tools and dashboards. On their own, these work fine. But as soon as they need to connect to existing systems, data or processes, there appears to be no foundation under them. The IT landscape fragments. Data gets scattered. Management becomes more complex. What started as a solution grows into a new problem.

On top of that, the disciplines needed to do this right – AI knowledge, architecture, automation and integration – are rarely on one team. The AI specialist doesn’t know the legacy. The architect doesn’t understand AI as more than a plug-in. Everyone has a piece of the puzzle. No one has the whole.

Four disciplines. Four conversations. And a pilot standing still.

This is not a technology problem. It is a structure problem.

What is Agentic Application Development?

Agentic Application Development (AAD) is an approach where AI agents do the heavy development work, while humans set the direction. No tool. No platform. A way of working that combines the speed of AI with the structure that enterprise software requires.

The difference is not in the technology. It’s in the order.

Step 1 – Start with the right problem

Every journey begins with Sprint 0: an exploration phase in which the real issue is defined before any choice is even made. What do you want to achieve? What problem are you solving? Is customization the right choice, or is an existing solution a better fit?

The biggest pitfall in software development is not how you build – it’s what you build.

Step 2 – Build from a proven foundation

You’re not starting from scratch. The foundation is already there: everything enterprise software needs – access management, integrations, security, notifications, documents – is already there, tested and ready to use. AI agents build new functionality on top of this foundation. Automatic quality checks ensure that everything meets enterprise standards.

Step 3 – Validate continuously

No big bang delivery. You continuously see working software – expanded step by step, with your feedback incorporated immediately. First production-worthy application: realistically within 4 to 8 weeks of kick-off.

Build the right thing. Build it right. Build it fast. In that order, never the other way around.

Our magic at work

What we did

How does AAD benefit your organization?

Deliver faster without sacrificing quality

Functionalities that used to take weeks are now realized in days. Not because steps are skipped, but because repetitive work is automated and the foundation is already in place.

Software that remains manageable as it grows

Applications built to established standards remain manageable, even when they grow into mission-critical systems. No subsequent fragmentation. No management costs that increase as the landscape becomes more complex.

You are and remain the owner

The software is all yours. Standard code in common technologies – no proprietary language, no platform dependencies. Any qualified development team can maintain and further develop the software. You are not dependent on any one party – including Ciphix.

No licensing fees that penalize your growth

Whether your application serves ten or ten thousand users – costs do not change based on user numbers.

IT and business closer together

As development becomes faster and more iterative, ideas can be tested faster. The distance between what the business wants and what gets built shrinks. Better decisions. More value from technology.

AAD is complementary, not a replacement

AAD is not the solution for everything. Standard software and existing platforms remain excellent for generic processes. If you work with platforms like Mendix, UiPath or Workato, you don’t have to choose. AAD fills the gap for those applications where full customization offers more value.

The question is not which approach is better. The question is which approach fits your particular issue.

AAD is the right choice if:

Your process is unique and standard software is not a good fit
you want full control over your software, without dependence on an external vendor
you want to grow without licensing costs growing with you
You have complex integrations with legacy systems that need to keep running
you want to use AI structurally – not as an experiment, but as part of your operation

When which approach?

Issue	Standard SaaS	Low-code / platform	AAD
Generic or standardized process	✓ Ideal	✓ Strong	– Overkill
Unique process, customization required	○ Limited	○ Sometimes possible	✓ Ideal
Complex integrations with legacy	○ Troublesome	○ Requires extra work	✓ Strong
Full ownership, no lock-in	○ Vendor-dependent	○ Platform-dependent	✓ Fully yours
Grow without increasing licensing costs	○ Per user	○ Per user/environment	✓ No per-user cost
Fast prototype validation	○ Slow	✓ Strong	○ Less capable

Organizations that made the move

Two examples of organizations that took an approach that works in production, not just in a demo.

Mental Care Group: frontrunner in healthcare

Mental Care Group was dealing with outdated systems and high workloads. Instead of waiting for a SaaS vendor, they took charge themselves. Together with Ciphix, they built a new registration, client and practitioner portal – fully integrated into their IT landscape. Result: a fundamentally different way of working in healthcare, and a position as a leader in their sector.

JBT Marel: more sales, not more people

JBT Marel wanted to grow in sales without growing in the number of employees. The solution: 65 digital employees performing daily work that was previously done manually. Only possible because the foundation – the integrations, the architecture, the governance – was set up right beforehand.

The questions you probably have

Aren't we dependent on you guys once this is built?

No. The software is all yours, standard code that any qualified development team can adopt. You are not dependent on Ciphix. If you quit the maintenance package, the application will continue to run.

How do we prevent the project from stalling?

Every project starts with Sprint 0, scope and priorities are established before construction begins. The iterative approach ensures that you continuously see working software and can make adjustments at any time.

Is AI-generated code reliable enough for mission-critical software?

Quality is not a matter of trust, it is a matter of control. All output is automatically checked for quality and safety before it reaches production. The speed is in eliminating repetitive work, not in skipping quality steps.

We already work with platforms. Should we replace them?

No. AAD does not replace your existing IT landscape, it complements it. Existing systems and platforms remain intact. Which applications benefit from AAD is up to you; we help focus that question.

How do we know we are building the right thing?

By starting with Sprint 0 – not with technology, but with the problem. What do you want to achieve? Is customization the right choice or is there a better way? Only when those questions are answered do we start building.

Transparent about cost and ownership

AAD is built around ownership, not dependency. The model is simple:

One-time project fee – you pay once for the build. You receive working software including full source code. No ongoing obligations for what is built.

Annual maintenance fee – access to the foundation that we are constantly developing: updates, security patches and new features. If you stop maintenance, the application continues to run and you receive a full source code copy.

Optional hosting – we provide hosting on EU infrastructure, or you deploy in your own cloud environment. Same application, your choice.

No per-user costs. Whether your application serves 10 or 10,000 users; costs do not change based on user numbers.

The organizations moving now are building a lead

AI is going to play a role in how your organization builds software. That role is already there. The question is how you fill that role, in a way that works in your organization, with your systems, at your scale.

AAD shows that speed and quality need not be mutually exclusive. Start with the right problem, build from a solid foundation, bring all disciplines together in one team. The result is software that is not only built fast, but also keeps working.

Ciphix rode out three technology waves for the market: low-code, RPA and iPaaS. In each wave, we helped organizations become leaders. Now we are taking the same step. But what is happening now with AI is of a completely different order.

The fourth wave is here. The winners are not the fastest builders. They are the organizations that know how to anchor it properly.

Ready to take the next step?

Every organization has a different starting point. We always start with Sprint 0 – an exploration phase in which we define the real issue together. No standard approach, but a focused exploration of where the greatest opportunity lies for your organization.