From client conversation to process diagram in hours: AI process modelling at Florinet.

Application: Process discovery and redesign of client processes with ModelMatic.

Florinet supplies an ERP package that helps growers in ornamental horticulture make their processes easier and clearer. With every new client, the team first has to get a sharp picture of how that process fits together. This used to be manual work: holding conversations, writing up notes, drawing a diagram and adjusting it. With AI process modeling through ModelMatic, that work has been cut from a full day to a couple of hours. "There is an enormous time saving in it," says Managing Director Wouter Jongkind. "Easily four or five hours per process we work out for a client."

The challenge: making every client process clear again

Florinet is growing fast, and that growth puts pressure on the one thing that slips first: documentation. For new clients the process has to be mapped from scratch, including where the current functionality pinches. For existing clients it becomes harder to keep the documentation up to date, while developers and the support department need exactly that context to place a question or an incident.

Drawing a process by hand takes time, and the result is only as good as the attention you have for it at that moment. Florinet was looking for a way to get to a reliable process model faster, without giving up detail or overview.

The approach: from interview to flowchart with ModelMatic

The way of working has stayed simple. Florinet records a client conversation or makes a report of it, and feeds that into ModelMatic. Out comes a structured flowchart that the team then walks through together with the client. The process discovery that used to take hours shifts to the real work: thinking about bottlenecks instead of drawing.

What Jongkind particularly values is that the process modeling pulls the different systems apart. In ornamental horticulture, Florinet connects to countless parties, from dashboards and the auction to systems on the client side, and the process diagram shows through separate swimlanes where each part takes place. "You see exactly which process happens at which system and what we are ultimately responsible for," he says. "I think that is really clever."

The result: hours saved per process

The biggest gain is in time. For a new client, working out a process saves easily four or five hours, with the same insight as before but without the manual drawing. For a redesign at an existing client, the saving is between two and eight hours, depending on how big the change is.

That redesign has become much faster since version two. Florinet can load the existing situation, apply the change to it, and build further on what is already there.

"You can load the old situation, add to it and keep building on it. That is an even bigger time saving, because otherwise you would have had to start from scratch."

Just as important is that the model is often right the first time. The corrections that remain are cosmetic, not substantive errors.

"What you put in and how it is translated is often spot on. Sometimes there are small cosmetic things, but that is purely layout."

A result you dare to show the client

The flowchart from ModelMatic is Florinet's calling card toward its clients. That it looks neat and clear helps in that conversation, especially with the option to export the process diagram as PowerPoint and present it more easily. Clients respond well to the result.

Looking ahead, Jongkind sees the biggest opportunity in collaboration, functionality he saw in the demos and that Florinet still wants to put to use. "Being able to work as a team, look over each other's shoulders and complement each other on projects, that seems ideal to me." For a growing team that wants to share knowledge about clients, that is the logical next step.

Why ModelMatic

For Florinet, AI process modeling does exactly what made process discovery, analysis and redesign heavy and time-consuming: the manual drawing. What remains is the substantive work: finding bottlenecks, making choices, taking the client along. Getting to a process model that is right, faster, with less back-and-forth and without starting from scratch every time.

Want to see how AI process modeling speeds up your process discovery? Book a demo.