AI is transforming global supply chains, delivering benefits that go beyond the initial hype. As a key driver of digitalization, AI enables faster, more adaptable supply chains and empowers customers with valuable data insights. Alex Nederlof, Director of Engineering Customs at Flexport, shares how AI is reshaping logistics—from customs automation to seamless document processing.

How would you describe the current state of AI in supply chains? Is it living up to the hype?

AI is at an exciting and transformative stage in supply chains. It’s not just a trend; it’s becoming a real game-changer. Right now, we’re seeing more companies—including some of the biggest names in the industry—embracing AI. About 65% of Fortune 500 companies mention it in their reports. However, over two-thirds also mention the potential risks, so it’s clear that companies need a solid strategy to harness AI’s full benefits while being mindful of the challenges.

At Flexport, we’ve been using AI in our products for over three years. What’s really new now is how much more capable and versatile AI has become with the rise of large language models (LLMs). Think about it this way: if you see AI deliver a result that’s just “okay” today, that result will only get better over time as the technology improves. Every year, it’s becoming significantly more accurate, cheaper, and faster, making it better at handling increasingly complex tasks. Plus, it can now work with new types of data, like images and audio. We’re already using these increasingly powerful models in our document digitization processes, which is a big deal for customs and procurement. In short, AI is evolving fast and will only continue to become more valuable for supply chains.

Could you give some specific examples of where AI can make an immediate impact?

One of the simplest but most impactful uses of AI is document processing. Traditionally, handling unstructured logistics documents—massive piles of shipping contracts, faxes, customs documents, and more—means copious manual work and back-and-forth. Within customs, for example, operators often spend a large portion of each day identifying missing data and documents, and then reaching out to individual customers to chase down the information in question.

With AI, we can automatically parse, understand, and organize these documents. What’s even better is that we can automatically find out if data is missing from documents, or if certain documents are missing altogether, and automatically follow up with the parties responsible for the data. This puts time-consuming back-and-forth on autopilot, significantly improves on-time performance for our customers, and makes it much easier to keep operations running smoothly.

And it doesn’t stop there. At Flexport, we’re taking it further and aiming to automate other cumbersome workflows. Our customs team is on track to automate 80% of manual tasks by 2025, such as filing out port routes and similar types of mundane paperwork. This means AI is helping us take on more data, handle more types of documents, and manage ambiguity better than ever before. So, in short, AI is speeding up processes, making data more accessible, and opening doors to automation we couldn’t even consider a few years ago.

Many companies struggle with the data foundation needed for AI. What does Flexport advise here?

Data is the fuel that powers AI. Without good data, even the most advanced AI models can’t deliver their full potential. We recommend that companies start by owning, organizing, and cleaning up their data. Work with vendors to ensure everything is recorded in a standardized, usable format that makes analysis and integration easier. This step is crucial, because having clean, accessible data makes all the difference when it comes to how well your AI initiatives will perform.

One of the great things about today’s AI models is their ability to handle data in all sorts of formats, including older, analog data. So, while managing data might seem messy, it becomes much more manageable with the right tools. We focus on structuring the chaos, turning even the most analog parts of the supply chain into a form that AI can understand and work with. This organized data foundation sets you up for success with any AI project.

Finally, how does Flexport recommend selecting tools for AI readiness?

One word: interoperability. That’s just a fancy way of saying that your tools should be able to “talk” to each other. We advise companies to pick tools that have strong API capabilities, so they can connect and share data easily across systems. This kind of connectivity allows AI insights to be shared with different teams, which means better, more informed decisions.

Today’s AI tools are also becoming more flexible, which helps with handling different data formats. For example, let’s say a customer sends us data in a format that doesn’t match what we need. Instead of spending hours manually reformatting it, we can use AI to handle that automatically. This saves a ton of time and effort, making it easier to keep things running smoothly without constant adjustments.

The future of AI in supply chains is incredibly promising. As companies keep testing and refining their AI applications, they will see the benefits grow. AI will become a major differentiator and help supply chains become more agile and resilient, especially in uncertain times.