From ‘digital native’ to ‘AI native’

AI is increasingly finding its way into our lives—including the greenhouse horticulture sector. As a result, data-driven operations are gradually becoming the norm. But how can you effectively integrate AI into your business operations? And where do you start? Digital freight forwarder Shypple shares how they transformed the organization and tracking of international freight forwarding into a fully AI-driven process, without compromising on personal service.

At GreenTech, ‘PURE’ represents clarity and purpose in crop production. It’s about skilled professionals working with care and responsibility, applying well-considered methods within controlled environments that ensure consistency, transparency, and trust. We spotlight what truly drives innovation in horticulture. In the GreenTech monthly newsletter series, we’ll explore ‘PURE’ from multiple perspectives, bringing you fresh insights, trends, and technologies shaping the future of growing.

After completing a Master’s degree in Finance and driven by a long-held desire to become an entrepreneur, Jarell Habets founded Shypple in 2016. The Dutch company organizes international sea and air freight via a digital platform, giving importers and exporters visibility into their shipments—from booking and documentation to real-time tracking and delivery.

“I had wanted to become an entrepreneur from a young age,” says Habets. “But I never really followed through on that plan—until I landed an entrepreneurial part-time job with a shipping freight forwarder during my master’s studies.” Habets likens the work of a freight forwarder to that of a travel agency for cargo. The company purchased cargo space on ships and aircrafts at various locations, and customers then bought this space to have their products transported from point A to point B. “That process was something of a ‘black box’ for customers; they were in the dark about the exact location of their cargo.”

Lack of visibility

Habets saw an opportunity to do things differently and founded Shypple. The name alludes to "shipping made simple." His ambition was to modernize the traditional logistics sector and provide the customer with full visibility into the entire process.
"The lack of transparency regarding the process was a pain point in logistics. Containers—especially those carrying fresh produce—arriving later than expected can result in significant losses. The advantage of starting a new company was that, as a 'digital native,' I could go 100% digital right from the start."
The first employees hired were software developers. Together, they built a platform that provides visibility into the entire journey. "Customers can see exactly where a shipment is located. All documentation is available online in a single place." This includes details such as port and terminal information, customs forms, inspection reports, and more. Transparency and predictability throughout the logistics chain are crucial, particularly when dealing with fresh produce.

AI works 24/7

If Habets could start over, he would build the company to be AI-native from the ground up. “It is always harder to transform an existing structure to accommodate new technologies. We are now viewing our current business through an AI lens: how can processes be organized more efficiently and quickly? A large part of our work involves communication, taking place via chat and email.”
This is where large language models (LLMs)—such as ChatGPT, Claude, and Gemini—offer a solution. Habets notes: “AI is far superior to humans when it comes to verifying information at scale. Moreover, it operates 24/7. That is a major advantage given the various time zones involved.”

AI can assist in resolving issues that have a clear right-or-wrong answer. A shipment arriving on time is good; a shipment arriving late is bad. Accuracy is crucial when it comes to information on forms; an incorrectly entered VAT amount, for instance, can cause significant delays. “We used to check all forms manually. By now, AI handles 90% of that process, with our teams double-checking the exceptions.”

Another example is predicting ship arrival times. “Ports and terminals possess accurate data on arrival times—precise information you wouldn't get from shipping lines. Here, too, AI can rapidly analyze vast amounts of data from terminals and ports.”  This information allows Shypple to improve expectation management. “If we see that a shipment of blueberries is arriving two days later than planned, we notify our customers. They can then take action—such as arranging an air freight shipment—to ensure they meet their delivery obligations.”

Boosting customer satisfaction

AI can also help analyze customer behavior. “We have been operating entirely digitally for ten years, which has generated a wealth of data. Using AI, we identified the most common complaints and analyzed why certain employees received higher satisfaction scores than others.”

For instance, Shypple discovered that communicating openly about bad news contributes to customer satisfaction. Habets cites the example of a delay caused by a storm in Hong Kong. “Customers appreciate it when we mention this and keep them updated daily—even when there is no new information.”

AI is changing the nature of work

Sufficient data forms the foundation of a good AI model. Habets notes: “If you implement AI effectively, it gives your organization a boost. Employees become much more productive, leaving more time for personal interaction with customers.” He expects artificial intelligence to transform the nature of work. “Some jobs will disappear—think of translation agencies, for example. Our platform operates in Chinese and Japanese as well, using a translation engine. Software development is moving much faster, and the demand for new features is endless.”

Habets cites a feature for reading emails using AI as an example. The development of this feature is also supported by AI, which reduces development time and increases returns. Shypple is currently focusing its hiring efforts on software developers, who are assisting in the transition to a fully AI-driven company.

Collecting data and experimenting

Collect data. That is Habets’s advice to companies looking to implement AI. “The more data you can store, the better. It doesn’t necessarily have to be structured. You can then use smart tools to analyze the data.” The second step is experimentation. “Come up with everything you’d like to achieve with AI, then create a top 10 list of ideas you genuinely want to test. After a few days of testing, one to three ideas will naturally emerge as the ones to pursue further. Free up the funds and capacity to develop them. The return on investment will follow.” Finally, Habets notes that even a failed test can be valuable. “At the very least, you’ll know not to look any further in that direction.”