From prompt to practice

AI is currently bringing about a fundamental change in the way software is developed. Requirements can be described, structures proposed and applications built in ever shorter times.

 

We are also consistently pursuing this approach with the flexmobility platform. However, our aim is not to demonstrate as spectacularly as possible what can be generated using a prompt alone.

We are interested in the more challenging question:

 

How can AI be used to create a business app that reliably maps a real-world business process?

That is precisely why our new AI features are now being put into practice with deliberately complex requirements.

AI can already be used to prepare large parts of an application. Fields, structures and functions can be derived from business specifications and incorporated into an application more quickly.

But a business is not made up of standard prompts.

 

A service technician works differently from a warehouse worker. A production process has different requirements to a sales process. ERP structures, authorisations, business rules, data quality and special cases vary from company to company.

This is precisely where human knowledge remains crucial.

 

They know their processes. They know why certain workflows have to function in exactly this way, what exceptions exist and what result is actually required in the end.

AI does not replace this knowledge. It makes it easier to develop a powerful business app based on it.

We therefore do not see AI as a replacement for the people behind an application.

Rather, a new division of tasks is emerging: AI can take on a large proportion of the preparatory and repetitive work. The crucial details remain where experience, an understanding of processes and company-specific knowledge are required.

 

The oft-mentioned final 20 per cent of a solution is by no means unimportant. On the contrary: it is precisely here that the difference often lies between an application that simply works and a business app that truly supports a business process effectively.

 

For us, this is precisely where a major opportunity lies.

If AI takes over a significant proportion of the work previously required, configurators can invest more time in these crucial details.

Less time on routine tasks. More time on the quality of the solution.

There is another point that is crucial to our approach.

At flexmobility platform, the AI does not start from scratch. It operates within an existing platform for native business apps.

 

This means that key foundations are already in place: the platform’s configuration logic, integration with enterprise systems and the capabilities of native applications.

The AI’s task is therefore not to build a separate technological ecosystem for every new application.

 

Instead, it is designed to help utilise the capabilities of the existing platform more intelligently and significantly more efficiently.

This is a key difference.

After all, an application must not only be developed quickly. It must also be possible to maintain, expand and adapt it to new requirements in the long term.

For our beta, we have deliberately chosen not to focus on particularly simple use cases. We have already tested these extensively in-house.

 

Several companies have expressed an interest in the new technology. Together, we are examining real-world processes with demanding requirements and varying conditions. This is precisely where we want to learn just how far the support already goes and where we can improve it further.

 

For us, therefore, a beta is not a perfectly staged product demo.

It is the moment when technology meets reality.

The experiences of the participating companies help us to further develop the AI support in line with actual requirements.

For businesses, this offers a particularly attractive long-term economic benefit.

If new requirements can be implemented in applications more quickly, project durations are reduced. Adjustments can be made with less effort, and existing specialist knowledge can be transferred more directly into digital processes.

 

At the same time, people remain an integral part of the process.

They define requirements, evaluate results, contribute their experience and decide when an application accurately reflects the actual business process.

 

Our goal is therefore not:

AI builds the application.

But rather:

AI and the platform provide people with tools that enable them to build high-quality applications much more quickly.

The launch of the beta is therefore an important step in our development.

We can now see, through real-world processes, how the interplay between AI, a powerful platform and human expertise works.

 

We are convinced that it is precisely this combination that can bring about lasting change in the creation of business apps.

 

Not by removing people from the development process.

But by relieving them of a significant portion of the work that currently prevents them from focusing on what really matters:

their process, their experience and the quality of their application.